Petitioner-Cross-Respondent Fluor Daniel, Inc. ("Fluor Daniel" or the "Company") appeals the decision of Respondent-Cross-Petitioner National Labor Relations Board ("NLRB" or the "Board") finding that Fluor Daniel violated Sections 8(a)(1) and (3) of the National Labor Relations Act (the "Act"), 29 U.S.C. §§ 158(a)(1) and (3) (1998), by failing to hire 124 applicants because of their union affiliations. The NLRB filed a cross-petition for enforcement of the unfair labor practice order issued by the NLRB after review of the case. Three unions, including two of the original charging parties, filed briefs as intervenors. Specifically, Fluor Daniel argues that: (1) the NLRB failed to require the General Counsel to allege and prove at the liability stage that jobs existed at the time the discriminatees had active applications on file; (2) the NLRB failed to take into account Fluor Daniel's hiring rules and policies regarding applicants when making its findings with respect to liability; and (3) the NLRB's order is not supported by substantial evidence.

For the reasons stated below, we find that the NLRB's decision that Fluor Daniel violated §§ 8(a)(1) and (3) of the Act1 was supported by substantial evidence and we GRANT the NLRB's cross-petition for enforcement for all employees except the five rebar helpers. We DENY the Board's cross-petition in regard to the rebar helpers and REMAND to the Board for further proceedings consistent with this opinion.

Fluor Daniel is the nonunion subsidiary of the Fluor Corporation, a holding company that also has a unionized subsidiary, Fluor Constructors. Fluor Daniel has a unionized subsidiary called TRS. Fluor Daniel, Fluor Corporation, and TRS are separate corporations with separate boards, officers, labor relations, and personnel policies.

In 1993, Fluor Daniel was hired for two distinct contracting projects — one to rebuild a coker plant destroyed by fire at the Exxon refinery in Baton Rouge, Louisiana ("Exxon"), and another to take over maintenance of the nuclear power plant at the Palo Verde Nuclear Generating Station near Phoenix, Arizona ("Palo Verde"). Fluor Daniel intended both projects to be "open shop," i.e., employees at the sites would not be part of labor organizations and there would be no union labor contract between Fluor Daniel and the hired employees.

Fluor Daniel has developed a hiring priority that it applies to all of its projects. The Company gives first preference to previous Fluor Daniel employees who have been certified through the Company's craft certification program. In order to be certified, Fluor Daniel employees are given a written test, or a demonstration test for welders, and an evaluation of their job performance after a certain period of time on the job, usually thirty days. Employee certification also requires a total of forty-two months of craft experience. Employees attaining certain scores on both their written or demonstration test, and job evaluation, are eligible for in-house certification in their craft. An employee may be certified in more than one craft.

When a project begins, Fluor Daniel first looks to its company database of employees with certification, and sends those former employees solicitations by mail, called mailgrams, describing the job and inviting them to the jobsite for processing, interviewing, and possible hiring. After attempting to fill open positions with certified applicants, Fluor Daniel prefers to hire applicants with previous Company experience. Lastly, Fluor Daniel looks at all other applicants, including those from the general public responding to newspaper advertisements and job postings at project sites. According to Company policy, all applications submitted at jobsites are entered into the Craft Employment Application Disposition Log and held active for sixty days from the time of submission. Additional Company policies adopted at jobsites include policies that applications were only accepted when positions were available and that applicants could only apply for one position.

Fluor Daniel was hired by the Arizona Public Service Company ("APS") to provide service and maintenance at Palo Verde, the largest nuclear power facility in the United States, for a total of three years beginning in the summer of 1994. Fluor Daniel was responsible for the maintenance of the plant generally, and for maintenance and refueling during scheduled outages when the plant was shut down. Before Fluor Daniel won the bid, the plant had been constructed, serviced and maintained by Bechtel Corporation ("Bechtel"), one of the world's largest general contractors and a unionized employer. In performing its contract at Palo Verde, Bechtel previously had contracted with craftsmen from several local unions, including the Boilermakers Local 627, Millwright Local 1914, and Ironworkers Local 75.

In its bid proposal for the Palo Verde project, Fluor Daniel stated that it had "determined that an open shop labor posture can best meet our goals." The proposal also emphasized that Fluor Daniel had a database of its former employees that included many with nuclear experience, and that the Company had the expertise to transition Palo Verde from a closed shop facility, i.e., one where all employees are union members, to an open shop facility.

Staffing at Palo Verde began in June 1994, and by February 1996 Fluor Daniel had hired 962 craft employees. Periodically, APS would prepare requisitions called Contract Labor Requests ("Requests") for Fluor Daniel, which indicated the number and type of craft employees that were needed at any particular time. Based on the Requests, Fluor Daniel was authorized to hire a certain number of employees for its base crew and for meeting its staffing needs during outages. Fluor Daniel relied heavily on telephone and mailgram recruiting of former employees, including many who lived outside of Arizona. The Company also posted available positions at the jobsite and allowed the general public to submit applications on a walk-in basis. Out of the 200 former Bechtel employees who previously had worked at the site, received training from the APS, and obtained security clearance, ninety-one were hired by Fluor Daniel.

Area trade unions decided to organize and apply for jobs at the site as voluntary union organizers ("VUOs"). Gary Evenson, a paid organizer from a local union of boilermakers, organized the effort at Palo Verde with former Bechtel employees. The VUOs "agreed to accept employment if offered, to stay until laid off, to do a good job ... to discuss the benefits of union representation with other employees [on the site] and to record notes of actions that appeared to interfere with protected rights." They were instructed to wear conspicuous union insignia and write "voluntary union organizer" on their applications. Fifty-two former Bechtel employees agreed to serve as VUOs and submitted applications at the jobsite between June 16 and June 23, 1994. None of these fifty-two former Bechtel employees were hired, though monthly reports in 1995 indicated Fluor Daniel had trouble meeting staffing needs for the project. On June 27, 1994, twenty-six former Bechtel employees serving as VUOs attempted to submit applications and were refused.

After winning the Exxon bid, Fluor Daniel recruiters solicited applications from former Fluor Daniel employees through the Company database by mailgram, called other jobsites that were closing, and maintained a telephone log of persons calling the Company for work at the site. Recruitment at the site was coordinated by Senior Site Manager Bill Austin and recruiter Rhonda Glover. Fluor Daniel also began three weeks of open staffing, posting job notices on-site and advertising in local papers. By January 19, 1994, Company recruiters had received 700 applications. Fluor Daniel followed its general hiring policies and procedures except that it made a significant change to the length of time for which applications were held active. The Company claimed that the urgent nature of the Exxon project necessitated a deviation from its normal sixty-day active application rule, and the implementation of a thirty-day active application rule. This deviation from Company policy was recommend by Ed Martinez, Industrial Relations Specialist, and approved by Jim Sayre, Regional Human Resources Manager at Fluor Daniel in Sugarland, Texas.

Area union organizers decided to organize an effort similar to that organized at Palo Verde. The Boilermakers union's business agent, Kendrick Russell, and an organizer of the Pipefitters' union, Jeff Armstrong, coordinated the effort. To that end, groups of VUOs submitted applications at the Exxon site on six occasions between January 19, 1994 and May 10, 1994. As in Palo Verde, applicants wore union insignia and indicated their union affiliation on their applications. None of these applicants were hired immediately, though one who had applied in January was hired after Glover was informed by Martinez that hiring union-affiliated applicants would improve the chances that Fluor Daniel could avoid a complaint from the NLRB. In total, Fluor Daniel hired nearly 2,800 employees by the end of the Exxon project in December 1994.

The Boilermakers International Union ("Boilermakers") filed charges with the NLRB against Fluor Daniel in the Phoenix, Arizona regional office alleging various violations of the Act. The Boilermakers and the Pipefitters Local ("Pipefitters") filed charges of various unfair labor practices against Fluor Daniel in the NLRB's regional office in Baton Rouge, Louisiana. The Regional Directors of the Board issued amended complaints in Phoenix and Baton Rouge on April 17, 1995 and March 29, 1995, respectively, alleging that Fluor Daniel violated §§ 8(a)(1) and (3) of the Act. On April 12, 1995, the General Counsel of the Board ordered consolidation of the charges. Administrative Law Judge Martin J. Linsky ("ALJ") held fifty-one days of hearings in Phoenix and Baton Rouge between August 1, 1995, and December 12, 1996. On February 6, 1998, the ALJ issued his decision, finding, most significantly, that Fluor Daniel had violated the Act by unlawfully discriminating against forty job applicants at Exxon and seventy-nine applicants at Palo Verde because of their union affiliation. Fluor Daniel was ordered to cease and desist from its illegal activity, hire the 119 applicants who had been refused employment or refused the opportunity to apply, to pay backpay and benefits to the applicants, and to post a notice at jobsites around Palo Verde and Exxon notifying the public of its violation of the Act.

All parties, including the charging parties, filed exceptions to the decision of the ALJ before the Board. The NLRB issued a Decision and Order on March 2, 2001, affirming the rulings, findings of fact, and conclusions of law of the ALJ. However, the NLRB modified the Order of the ALJ. In addition to several small modifications, the NLRB's decision made three significant modifications. First, the NLRB required Fluor Daniel to post the notice at all of its jobsites in the nation, not just in the areas surrounding the Exxon and Palo Verde sites, and to mail the notice to all employees. Second, the NLRB acknowledged that the ALJ inadvertently omitted a testing remedy and included a provision in its order that the discriminatees should be tested in order to receive Fluor Daniel certification. Third, the NLRB found that Fluor Daniel unlawfully refused to consider five additional applicants for rebar helper positions at Exxon (these positions involved assisting with laying the cement foundation of the coker) and ordered remedies for those individuals consistent with the remedies set forth for the other violations. On March 12, 2001, Fluor Daniel filed a petition for review of the NLRB's decision to this Court. The Board filed its application for enforcement on March 29, 2001.

The scope of our review of determinations by the Board is limited. We review factual findings of the NLRB to determine if they are "supported by substantial evidence on the record considered as a whole." W.F. Bolin Co. v. NLRB, 70 F.3d 863, 870 (6th Cir.1995); see also 29 U.S.C. § 160(e) (1998) ("The findings of the Board with respect to questions of fact if supported by substantial evidence on the record considered as a whole shall be conclusive."). "Substantial evidence" is "such relevant evidence as a reasonable mind might accept as adequate to support a conclusion." Lee v. NLRB, 325 F.3d 749, 754 (6th Cir.2003) (quoting Dupont Dow Elastomers, L.L.C. v. NLRB, 296 F.3d 495, 500 (6th Cir.2002)). We also review the Board's application of the law to the facts under the substantial evidence standard. FiveCAP, Inc. v. NLRB, 294 F.3d 768, 776 (6th Cir.2002). We may not displace the Board's reasonable inferences even if we would have reached a different conclusion. Id. Additionally, we severely limit our review of credibility determinations and accept those made by the Board unless they have "no rational basis." NLRB v. Valley Plaza, Inc., 715 F.2d 237, 242 (6th Cir. 1983).

Fluor Daniel argues that the NLRB failed to require the General Counsel to prove that openings existed for each discriminatee at the time they had applications pending, as required by this Court's decision in NLRB v. Fluor Daniel, 161 F.3d 953 (1998) ("Fluor Daniel II"). We decline to dismiss the charges against Fluor Daniel based on this argument. If the NLRB applies the correct standard of law, "[t]he Board's application of law to facts is also reviewed under the substantial evidence standard." NLRB v. St. Francis Healthcare Ctr., 212 F.3d 945, 952 (6th Cir.2000). There is substantial evidence in the record as a whole to uphold the finding of the NLRB that openings existed during the time that discriminatees submitted, or attempted to submit, applications.

The General Counsel bears the burden of proof in unfair labor practices cases. 29 U.S.C. § 160(c) (1998) (stating that violations of the Act can be adjudicated only "upon the preponderance of the testimony taken" by the NLRB); see also Section 7(c) of the Administrative Procedure Act, 5 U.S.C. § 556(d) (1996) ("Except as otherwise provided by statute, the proponent of a rule or order has the burden of proof."). This Court in Fluor Daniel II clearly articulated the proper burdens of proof in refusal-to-hire cases. The Court relied on the NLRB's decision in Wright Line, 251 N.L.R.B. 1083 (1980), enforced, 662 F.2d 899, 904 (1st Cir.1981), cert. denied 455 U.S. 989, 102 S. Ct. 1612, 71 L. Ed. 2d 848 (1982), that was approved by the Supreme Court in NLRB v. Transportation Management Corp., 462 U.S. 393, 402-03, 103 S. Ct. 2469, 76 L. Ed. 2d 667 (1983). In Fluor Daniel II, we adopted the standard set forth by the Supreme Court in Transportation Management, and held that "there are two elements to a Section 8(a)(3) violation: [ (1) ] anti-union animus and [ (2) ] the occurrence of a covered action — for example, a particular discharge, or a particular failure to hire." 161 F.3d at 966. We specifically rejected the NLRB's position that a simple showing of anti-union animus and the fact that certain individuals were not hired were sufficient to show a violation of §§ 8(a)(1) and (3), holding instead that an actual "failure to hire" had to be shown. Id. at 967. We specifically held that in order to make out a prima facie case for a refusal-to-hire violation, the General Counsel must first establish the two elements articulated in Transportation Management. Id. After the General Counsel has proven both elements, the employer must present evidence that the employees in question would not have been hired, even if they had not been involved with a union. Id. at 966. This analysis requires the General Counsel to "match[ ] up applicants with available jobs for which they are qualified...." Id. at 968. This means that employees involved in a violation of § 8(a)(3) must be actually qualified for the respective job positions and that the job positions must be actually available. Once the General Counsel makes a showing that the employer did not hire the qualified applicant for an available position, the General Counsel must also then show that the decision was motivated by anti-union animus. Id. at 966.

In light of our decision in Fluor Daniel II, the NLRB decided FES (a Division of Thermo Power) & Plumbers & Pipefitters Local 520 of the United Association, 331 N.L.R.B. No. 20, 2000 WL 627640 (May 11, 2000), to "give guidance to all parties litigating refusal-to-hire and refusal-to-consider violations by making clear the elements of the violation, the respective burdens of the parties, and the stage at which issues are to be litigated." FES, 2000 WL 627640, at *4. In FES, the NLRB held that in a refusal-to-hire case, the General Counsel bears the burden of showing:

(1) that the respondent was hiring, or had concrete plans to hire, at the time of the alleged unlawful conduct; (2) that the applicants had experience or training relevant to the announced or generally known requirements of the positions for hire, or in the alternative, that the employer has not adhered uniformly to such requirements, or that the requirements were themselves pretextual or were applied as a pretext for discrimination; and (3) that antiunion animus contributed to the decision not to hire the applicants. Once this is established, the burden will shift to the respondent to show that it would not have hired the applicants even in the absence of their union activity or affiliation. If the respondent asserts that the applicants were not qualified for the positions it was filling, it is the respondent's burden to show ... that they did not possess the specific qualifications the position required or that others (who were hired) had superior qualifications, and that it would not have hired them for that reason even in the absence of their union support or activity.

FES, 2000 WL 627640, at *6 (footnotes omitted)

Though the standard in FES could be interpreted as requiring a slightly lesser showing than that we articulated in Fluor Daniel II, the NLRB in FES fully addressed our concerns and set out a framework wholly consistent with our holding in Fluor Daniel II. The two decisions require that the General Counsel make specific findings both that jobs were available at the time of the alleged discrimination and that discriminatees were qualified for the jobs. Compare Fluor Daniel II, 161 F.3d at 968 with FES, 2000 WL 627640, at *6.

In the present case, the NLRB specifically acknowledged and applied the standards of FES in its analysis. Fluor Daniel, 2001 NLRB LEXIS 131, at *56 (Mar. 2, 2001). The NLRB articulated in its opinion the standards set out in FES and Fluor Daniel II and adopted the findings of the ALJ that the General Counsel had "clearly established his prima facie case." Id. With respect to the specific issue of job matching, the NLRB adopted the findings of the ALJ that Fluor Daniel was hiring during the time that the discriminatees applied and that the discriminatees were qualified. Id. In fact, the NLRB noted, all the discriminatees at the Palo Verde site were actually former employees of Bechtel and obviously qualified to perform the work. The NLRB rejected Fluor Daniel's challenge, raised in its exceptions, that the ALJ failed to match properly each discriminatee with an available job that they were qualified to perform, as required by this Court in Fluor Daniel II. The NLRB found that

[i]n contrast to Fluor Daniel II, in the proceeding before us, the administrative law judge expressly considered job availability and made factual findings that vacancies existed that the discriminatees were qualified to fill, and, further, that over the life of both projects there were enough positions to have employed every discriminatee.

Fluor Daniel, 2001 NLRB LEXIS, at *69. The NLRB found that the analysis undertaken by the ALJ comported both with our decision in Fluor Daniel II and the NLRB's decision in FES.

We agree. The ALJ undertook a detailed analysis of the discrimination at each of the work sites, noting names of applicants, the dates that they applied for positions or attempted to apply for positions, the number of applicants actually hired by Fluor Daniel, and the positions that were filled by Fluor Daniel. Fluor Daniel, Inc., 1998 NLRB LEXIS 1009 (Feb. 6, 1998). With regard to the Exxon site, the ALJ found that the "record is clear that the union affiliated applicants for employment applied for work and were qualified...." Id. at *39. The ALJ found that on five different days, union activists applied for employment and were not hired, and that on one day, four union pipefitters were not allowed to apply or to be tested. Id. at *50-*51. The ALJ continued his analysis and found that Fluor Daniel hired other applicants for the same positions and that Fluor Daniel had such desperate needs for electricians, pipefitters, and welders in August and September 1994 that it took fairly drastic action. Id. First, in August and September 1994, Fluor Daniel subcontracted with several nonunion firms in order to fill its staffing needs, and second, it sent out a total of 20,218 mailgrams to former employees in September and October 1994 looking for pipefitters and welders. Id. at *51-*52. Furthermore, regarding the discriminatees' qualifications for the job, the ALJ found that the applicants at Exxon were "highly qualified based on the material contained in their applications in the record." Id. at *35; see also Glenn's Trucking Co. v. NLRB, 298 F.3d 502 (2002) (finding that after examining the applications of individuals hired by the company, it was established that the discriminatees' credentials were in the same range).

At Palo Verde, the ALJ found that groups of discriminatees went to the jobsite and filled out applications on four different days in June 1994 and that one group of discriminatees attempted to submit applications but were refused on another day in June. See Fluor Daniel, Inc., 1998 NLRB LEXIS, at *15-*20. The ALJ "next inquir[ed] ... whether there were jobs that were to be filled." Id. at *24. The ALJ found that from June 1994 until February 1996, Fluor Daniel hired 962 craftsmen, that the applications submitted by the discriminatees were active, and that all the applicants had the required experience because they worked at Palo Verde previously for Bechtel. Id. at *25.

This is precisely the type of detailed analysis required under our decision in Fluor Daniel II, as incorporated by the NLRB in FES. Because the findings and conclusion of the NLRB are supported by substantial evidence on the record, we decline to disturb the NLRB's application of the law to these facts.

Fluor Daniel argues that the NLRB failed to take into account the hiring rules and procedures developed by the Company in its finding of liability. Specifically, Fluor Daniel argues that two rules should have been taken into account by the NLRB: (1) the rule stating that applications are active for sixty days; and (2) the hiring preference system.

Fluor Daniel's argument misunderstands the findings of the NLRB regarding its hiring rules and policies. The Company is correct that neither the ALJ nor the NLRB challenged the lawfulness of its hiring rules and policies per se. The ALJ concluded that, "I do not find on the basis of the record before me that [Fluor Daniel's] hiring practice per se is violative of the Act." Fluor Daniel, 1998 NLRB LEXIS, at *8. The NLRB declined review of Fluor Daniel's hiring practices, stating that "[b]ecause we find that [Fluor Daniel] applied its hiring policies and preferences in an unlawful manner, we find it unnecessary to reach Charging Party Boilermakers' argument that [Fluor Daniel's] hiring policies were inherently destructive of protected rights, and invalid per se." Fluor Daniel, 2001 NLRB LEXIS, at *10 n. 9. Instead, the NLRB found that Fluor Daniel had violated the Act because it "unlawfully applied its system of hiring preferences, policies, and procedures so as to refuse to consider or hire 120 voluntary union organizers...." Id. at *10.

1. The thirty-day active application rule at Exxon

The finding by the NLRB that the thirty-day rule was applied in a discriminatory fashion is supported by substantial evidence, and Fluor Daniel acknowledges that it deviated from its hiring policy in twenty-seven cases. Fluor Daniel instituted a thirty-day active application rule at Exxon in order to meet the urgent needs of the project. The NLRB did not address the lawfulness of this change in corporate policy, but found simply that the rule, though "enforced more rigorously than other rules," id. at *29, was applied in a discriminatory manner against the union applicants. Specifically, the NLRB found that Fluor Daniel failed to advise the discriminatees that the period for which their application would be valid had been changed. Id. The NLRB also found that the applications themselves lacked such notification, containing only a small line at the bottom of the page stating that the applications would be active for sixty days. However, the NLRB found that "[w]hen the alleged discriminatees' applications were submitted into evidence at trial, the '60' had been crossed out and '30' had been substituted by hand." Id. Furthermore, the NLRB found that the discriminatees testified that the forms they completed were not modified in any such manner, and there was no evidence presented that the applications of the nonactivists had been altered. Id. In contrast to the treatment of the activists on the Exxon project, the NLRB found that Fluor Daniel deviated from the thirty-day rule by hiring twenty-seven craft employees more than thirty days after they applied. Id.

Fluor Daniel, relying on Kelly Construction of Indiana, Inc., 333 N.L.R.B. No. 148, slip op. at 1 (2001), argues that the deviations for twenty-seven journeymen out of the 2,800 total persons hired is de minimis. This argument is without merit. The NLRB in Kelly held that an isolated and marginal deviation from a neutral, uniformly applied hiring policy would not constitute a violation of the Act. Id. Kelly does not support Fluor Daniel's conduct here because the thirty-day rule was not uniformly applied and because deviations from the rule can be considered neither isolated nor marginal.

Furthermore, the NLRB's discussion of the thirty-day rule was prompted by Fluor Daniel's assertion that its hiring policy constituted a defense to the allegations that it had violated the Act. Fluor Daniel sought to use the thirty-day rule as a reason for not hiring the alleged discriminatees. In considering this defense against the evidence presented, the NLRB concluded that the policy as applied to the union activists was pretextual. We conclude that the factual findings of the NLRB on this issue are supported by substantial evidence. Cf. Architectural Glass & Metal Co. v. NLRB, 107 F.3d 426, 432 (holding that a NLRB determination of the validity of a policy asserted as a nondiscriminatory reason for refusing to hire an applicant, is a question of law that is reviewed de novo). The NLRB found evidence from the Company's own files demonstrating both that Fluor Daniel had serious difficulty staffing the Palo Verde project and that the thirty-day rule was abandoned in favor of nonactivist applicants and rigidly applied against the discriminatees.

Fluor Daniel argues that the NLRB failed to take into account the Company's written policies regarding hiring preferences. As explained previously, Fluor Daniel gave first priority to applicants who were certified by the Company, second priority to applicants who had previously worked with Fluor Daniel, and third priority to all other applicants. Fluor Daniel argues that given these priorities, the length of craft experience required and the policy that applications were only active for sixty days, only twelve of the discriminatees at Palo Verde could have been hired absent Fluor Daniel's violation of the Act.

There is substantial evidence in the record to support the NLRB's finding that Fluor Daniels's hiring system was used in a discriminatory manner. We must give deference to the inferences drawn by the Board and must determine whether those inferences are reasonable in light of the facts. See, e.g., NLRB v. Ky. May Coal Co., 89 F.3d 1235, 1242 (6th Cir.1996). Fluor Daniel is correct in asserting that the NLRB never challenged the validity of hiring preferences per se. But, as with the thirty-day application rule, the NLRB found that Fluor Daniel applied its preferences in a manner which discriminated against the union activists. In particular, the NLRB found evidence that the scope of the preferences was expanded to include all former Fluor Daniel employees (not solely certified ones), without proper authorization. Fluor Daniel, 2001 NLRB LEXIS, at *61-*62.

Regarding the Exxon site, the NLRB found that since "[n]early one-fifth of Exxon journeyman hires were non-preferenced," id. at *25, the use of the preference system could not explain why the discriminatees were not hired. The NLRB also found that Fluor Daniel hired forty-three former employees in lieu of discriminatees, even though those former employees lacked the required forty-two months of craft experience, were certified in crafts other than those for which they were hired, whose applications had expired, or were considered for crafts other than those for which they had applied. Id. at *31-*34. Furthermore, the NLRB found that Fluor Daniel's proffered defense that the discriminatees did not have the necessary forty-two months of craft experience was also pretextual. Id. The NLRB found that many of the discriminatees had twenty to thirty years of experience; that Fluor Daniel recruiters never informed them that their applications contained insufficient experience; that Fluor Daniel did not give credit for training in union apprenticeship programs, while it did give credit for training at Fluor Daniel's craft school; and that over a hundred other applicants had less than forty-two months of experience. Id. In sum, the NLRB concluded that Fluor Daniel used its hiring preferences as a pretext for discriminating against the union activists. A review of the record shows that the inference made by the NLRB on this issue is supported by substantial evidence.

D. The NLRB's Decision and Order is Supported by Substantial Evidence

In its final assignment of error, Fluor Daniel argues that the Decision and Order of the NLRB is not supported by substantial evidence. The Company argues that "the Board ignored a significant part of the evidence and drew unreasonable and illogical inferences and conclusions from the evidence it did consider." Specifically, Fluor Daniel makes four arguments relating to errors in the findings of fact and subsequent conclusions of the Board: (1) the Board erroneously credited testimony by Gary Evenson, a witness for the Charging Party Boilermakers; (2) the Board erroneously substituted its judgment for the business judgment of Fluor Daniel; (3) the Board erroneously concluded that Fluor Daniel considers only nonunion employees to be loyal; and (4) the Board erroneously found that Fluor Daniel had large numbers of deviations from its hiring protocols. All of these arguments are without merit, as substantial evidence supports the findings of the NLRB.

First, Fluor Daniel asks us to reassess the ALJ's credibility determination concerning Evenson, a key witness from Palo Verde, based on the fact that Evenson was found to have filed frivolous unfair labor practice charges, falsified evidence, and committed perjury in a 1993 organizing event in the case of Irwin Industries, Inc., 1996 NLRB LEXIS 666 (Sept. 18, 1996). However, we may not review credibility determinations unless they have no rational basis. Valley Plaza, Inc., 715 F.2d at 242. The ALJ was under no obligation to consider determinations made by another ALJ in a wholly different case regarding the credibility of a particular witness. During the hearing, the ALJ ruled under Rule 103 of the Federal Rules of Evidence that any testimony regarding Evenson's involvement in Irwin Industries would add confusion to the current case. Furthermore, in Irwin Industries, 325 N.L.R.B. 796, 797 n. 7 (1998), the ALJ was reversed by the NLRB, who explicitly stated that "we do not agree with the [ALJ] that the unfair labor practice charges related ... to the three employees [including Evenson] were frivolous and knowingly false." After hearing the testimony and observing Evenson at the hearing in this case, the ALJ concluded that his testimony was credible. There is no evidence presented to refute that conclusion or that should lead this Court to disturb the ALJ's credibility determination.

Second, Fluor Daniel argues that the Board erroneously substituted its judgment for Fluor Daniel's business judgment by suggesting what protocol should have been used for staffing the two projects. In particular, the Company argues that it was the APS, and not the Company, that determined staffing requirements for Palo Verde and that the Board penalized Fluor Daniel for not making certain business decisions, such as reviewing the Bechtel workforce for qualified employees.

Fluor Daniel's arguments mischaracterize the findings of the NLRB. In its Decision and Order, the NLRB never challenged the lawfulness of Fluor Daniel's systems of hiring preferences, policies, and procedures per se. Instead, the NLRB concluded that Fluor Daniel had applied those preferences, policies and procedures in a discriminatory way, and substantial evidence supports this conclusion. The NLRB noted several instances in which Fluor Daniel made choices that were contrary to the provisions of the Act and we review those factual findings only to determine whether they are supported by substantial evidence. See W.F. Bolin Co., 70 F.3d at 870.

The findings of the NLRB do not represent an attempt by the Board to substitute its judgment for the business judgment of the Company. Instead, the NLRB cited many instances of disparate treatment between the discriminatees and other applicants, noting in each case not that Fluor Daniel should have conducted business differently, but rather how Fluor Daniel treated discriminatees in a manner different from other applicants. For example, the NLRB found that Fluor Daniel discriminated against applicants in Palo Verde by failing to consider the applications of former employees of Bechtel, an obviously qualified and available pool of potential employees, while actively seeking applications from other sources. The NLRB found other instances where Fluor Daniel acted in contravention of the Act in Palo Verde by treating discriminatees differently: Fluor Daniel allowed several discriminatees to apply for nonexistent job classifications, while allowing other applicants to apply for comparable work with different job classifications; Fluor Daniel invited nonunion applicants to check back in the case of no-shows but did not invite the alleged discriminatees to do the same; and Fluor Daniel never referred any of the discriminatees to its Wolf Creek nuclear project, as it did for nonunion applicants.

Fluor Daniel's third and fourth arguments relate purely to findings of fact and inferences and do not warrant lengthy discussion, as we review them only to determine if they are supported by substantial evidence. See Id. Fluor Daniel argues that the NLRB erred when it found that Fluor Daniel considered only nonunion employees to be loyal. The Board found substantial evidence that Fluor Daniel harbored anti-union animus, in particular finding that Fluor Daniel did not consider union organizers to be loyal. Fluor Daniel, 2001 NLRB LEXIS, at *59. In support of this conclusion, the Board found that Fluor Daniel promotes its database of "loyal" employees to customers. Id. Indeed, in its proposal for Palo Verde, Fluor Daniel stated that it had "determined that an open shop labor posture can best meet [Palo Verde's] goals and provide them with `[w]orkers loyal to [Palo Verde] and Fluor Daniel instead of a local union.'" Therefore, the conclusions of the NLRB that Fluor Daniel openly expressed the view that union organizers could not be loyal to the Company is supported by substantial evidence.

Finally, Fluor Daniel argues that the NLRB erroneously found that it had a large number of deviations from its hiring protocol. In particular, at Exxon the NLRB found 3,000 deviations from Fluor Daniel's hiring protocols with respect to the hire of nonactivists, but strict adherence to several hiring protocols to exclude the discriminatees from consideration. The NLRB found deviations from the protocol with respect to three major areas — hiring of former employees, hiring of applicants with inactive applications, and hiring of applicants with insufficient experience. Id. at *35. As previously discussed, Fluor Daniel hired former employees (also referred to as preferenced employees) who did not have forty-two months of craft experience, who had been certified in crafts other than the ones that they were hired to perform, who had never submitted applications or whose applications had expired, or who were considered for crafts other than those for which they applied. Second, Fluor Daniel also hired twenty-seven journeymen in breach of the Company's thirty-day active application rule that it implemented for the Exxon project. Id. at *29. Third, Fluor Daniel hired forty-three former employees without the requisite forty-two months of experience, and 124 applicants without former company experience and with less than forty-two months experience at Exxon. Id. at *34.

The NLRB additionally found that at Exxon, contrary to the requirement that applicants submit only one application and seek consideration for only one position, Fluor Daniel considered two-thirds of all applicants for positions other than those indicated on their applications. Id. at *35. None of the discriminatees were afforded the same opportunity. The NLRB, after extensively reviewing the enormous amount of Fluor Daniel's employment data, made several other findings, including that: (1) of the applicants hired prior to submitting their applications, two-thirds had inadequate experience and two-fifths were considered for different crafts; (2) of the applicants hired more than thirty days after submitting applications, three-quarters had inadequate experience and four-fifths were considered for different crafts; and (3) of the noncertified former employees with less than forty-two months experience, two-thirds were considered for other crafts. Id.

The NLRB found 892 deviations from Fluor Daniel's hiring protocol at Palo Verde, all in favor of nonunion activists. Id. at *52. Analyzing the Company's own data for nonpreferenced hires through November 1, 1994, the NLRB found: nine journeymen were hired before they applied; two journeymen were hired after their applications had expired; eighty-eight journeymen were hired with less than forty-two months of craft experience; and over 700 applicants were considered for crafts other than those for which they applied. Id. The NLRB also found that "significant percentages of applications contained more than one deviation." Id.

The deviations found by the NLRB represent its factual findings after review of the evidence presented, including numerous documents and hearing testimony. Fluor Daniel argues that "the exhibits [presented] show the silliness of the Board's finding," because many of the deviations in which the craft on the applications differed from the craft hired were the result of spelling errors, the use of abbreviations and a vernacular substitute. Under our standard of review of factual findings made by the NLRB, we find that these factual finding are supported by substantial evidence. See W.F. Bolin Co., 70 F.3d at 870. The conclusions of the NLRB were the result of the examination and statistical analysis of more than 1,000 applications and accompanying Fluor Daniel databases of former employees and hires. Clearly, the findings of fact are supported by substantial evidence on the record considered as a whole, and Fluor Daniel's argument that some of the nearly 900 findings of deviations were the product of spelling errors or misunderstandings is simply unpersuasive. Given that the NLRB examined over 1,000 applications and found almost 900 deviations, even if some portion of these deviations were the result of misunderstandings, the record nevertheless contains substantial evidence to support the finding of the NLRB.

Finally, we hold that the standard set forth in FES for "refusal-to-consider" claims, as applied in this case by the NLRB to the five union members who sought positions as rebar helpers, is in conflict with our caselaw. Therefore, we remand this issue to the NLRB for further proceedings consistent with this opinion.

On April 19, 1994, five union members submitted applications for rebar helper positions in response to a job posting at the Exxon site. Both the recruiter and supervisor at the site acknowledged that postings were made only if jobs were available and a requisition had been completed. After the applications were submitted, the job posting was removed and the applications were never considered. The ALJ dismissed the failure-to-hire allegations against Fluor Daniel regarding these employees, apparently finding not only that no rebar helpers were hired after the employees submitted their applications, but that there were no jobs available when the union members applied. However, the NLRB concluded that the ALJ did not fully address the failure to-consider claim for these applicants.

The NLRB analyzed the failure to consider claim under the standard it set forth in FES. In FES, the NLRB reiterated its long-held position that "hiring need not take place in order to find an unlawful refusal to consider union applicants for employment." 2000 WL 627640, at *11. The Board explained that in order to find a violation in refusal to consider cases:

[T]he General Counsel bears the burden of showing ... (1) that the respondent excluded applicants from a hiring process; and (2) that antiunion animus contributed to the decision not to consider the applicants for employment. Once this is established, the burden will shift to the respondent to show that it would not have considered the applicants even in the absence of their union activity or affiliation.

Id. at *10. In adhering to this test in its analysis of the alleged violations of the Act by FES, the Board stated that it was unpersuaded by our position in Fluor Daniel II that there can be no violation of Section 8(a)(3) when no hiring is occurring. Id.

In Fluor Daniel II, we specifically held that violations of Section 8(a)(3) can only occur when an employer is hiring for the position(s) at issue. We noted that in order to prove a violation of Section 8(a)(3), the General Counsel needs to show (1) that there is anti-union animus, and (2) the occurrence of a covered activity, for example a particular discharge or a particular failure to hire. Fluor Daniel II, 161 F.3d at 966. Once the General Counsel proves these elements, the employer must show that the employees in question would not have been hired or considered even in the absence of their union affiliation. Id. In dicta we noted that: "An employer who has no need for a particular applicant's services has not failed or refused to consider an applicant in violation of the Act." Id. at 967 n. 15. We also stated that, "There is no interference with, restraint, or coercion of applicants in the exercise of their protected rights when an employer, even with anti-union animus, rejects applicants who are in fact unqualified or for whose particular services the employer simply has no need." Id. at 967.

We do not agree with the Board's use of the test set forth in FES and rejected by this court in Fluor Daniel II. We therefore remand this issue to the NLRB for a determination of whether in fact, as required for a violation of Section 8(a)(3) of the Act, there were jobs available for rebar helpers when the union members applied.

For the reasons stated above, we find the NLRB's conclusions supported by substantial evidence and we GRANT the NLRB's cross-petition for enforcement for all employees except the five rebar helpers. We DENY the Board's cross-petition in regard to the rebar helpers and REMAND to the Board for further proceedings consistent with this opinion.