Sears, Roebuck & Company, Petitioner-Cross-Respondent v. National Labor Relations Board, Respondent-Cross-Petitioner

MINER, Circuit Judge: Petitioner-cross-respondent Sears Roebuck & Company seeks review of a National Labor Relations Board (“NLRB”) order directing Sears to bargain with the union certified as the representative of Sears’ automotive employees at its Maplewood, Minnesota store, after a certification election. The NLRB found that Sears had committed an unfair labor practice in violation of section 8(a)(5) and (1) of the National Labor Relations Act, 29 U.S.C. § 158(a)(5) and (1), by refusing to bargain with the union. The NLRB cross-petitions for enforcement of its order. Sears contends that its refusal to bargain was justified because the NLRB Regional Director conducted an invalid certification election. Specifically, Sears claims that the election occurred without a prior determination of the size and composition of the bargaining units, thereby denying voters an opportunity to make an informed choice. In addition, Sears contends that the notice sent by the Regional Director to explain the situation regarding the undefined units failed to provide the necessary information. We hold that postponement until after the election of a determination as to the unit’s configuration was within the scope of the NLRB’s authority to prescribe election procedures, and did not violate the employees’ right to informed voting. Accordingly, we grant the Board's application for enforcement. BACKGROUND On January 29, 1990, the United Steelworkers of America, AFL-CIO, CLC (the “union”) petitioned for certification as the exclusive bargaining representative of the employees at the Sears store located in the Maplewood Mall in Maplewood, Minnesota. The store consists of two distinct areas. The first, the main sales floor, is attached directly to the shopping mall and houses most of the merchandise and about 320 employees. The second area employs approximately fifty workers and is known as the automotive “backshop,” where customer cars are serviced. The backshop is connected to the main Sears facility but not directly to the rest of the mall. The back-shop employs several different types of workers, including mechanics, service ad-visors, parts attendants, and cashiers. Each of these groups of employees performs specific functions when a customer brings a car in for service. Another group of five automotive-oriented employees is employed in the mall portion of the store. These “automotive floor sales employees” sell the company’s automotive products from the main sales floor. They do not service customer vehicles, but merely sell parts and accessories and advise customers on the proper usage of these products. As such, the automotive floor employees perform duties similar to those performed by members of the back-shop staff, particularly the service advisors and parts attendants. The union proposed the certification election for several different bargaining units comprised of various groups of employees at the store. One proposed unit consisted of the backshop staff in combination with the automotive floor sales employees. Sears supported this position. The Regional Director initially rejected this idea and split the store into two units, one composed solely of backshop employees and the other composed of all remaining employees. The company moved to reopen the proceeding for inclusion of the automotive floor sales employees in the backshop unit. The Regional Director denied the motion to reopen, but was unable to conclude which unit should embrace the automotive floor *55 staff. As a consequence, the Director ordered those employees to vote challenged ballots in both elections. Under this procedure, the automotive floor staff votes would not be counted with either unit immediately. If, after the final tabulation, their ballots were determinative of the outcome in either unit, appropriate unit placement for the automotive floor employees would be decided at a post-election proceeding. Sears sought review of this ruling. In the interim, elections were scheduled for June 1, 1990. On May 31, the NLRB denied Sears’ request for review. In preparation for the elections, the Regional Director sent a notice for Sears to post in the workplace explaining how the election would be conducted. The notice consisted of three panels, with an attachment folded up behind the middle panel for mailing purposes. The attachment detailed the special voting posture of the automotive floor sales employees and the circumstances for including their votes. Sears posted the notices several weeks before the election, but failed to fold down the attachment. Approximately twenty minutes before voting commenced, the monitoring officer from the Regional Director’s office realized that the attachment had not been exposed and corrected the mistake. Thus, not all the voters were notified as to how the automotive floor staff votes would be counted. The union lost the certification election by a large margin in the unit that included all but the backshop employees. In the unit composed solely of backshop employees, however, the union prevailed by a five vote margin, twenty-four votes to nineteen, causing the as yet uncounted ballots of the five automotive floor sales employees to be determinative. The Regional Director’s subsequent investigation led to a decision to place the automotive floor staff in the unit with the backshop personnel. The additional ballots of the floor staff then were counted, and the union thereupon prevailed by a margin of twenty-five votes to twenty-three. The NLRB certified the union as the unit’s exclusive bargaining representative, but Sears refused to bargain with it. The union filed an unfair labor practice charge, and the Regional Director issued a complaint. The company answered by denying the validity of the election and hence the NLRB’s certification of the union. The NLRB granted the General Counsel’s request for summary judgment and directed Sears to bargain with the union. Sears then petitioned this Court for review of the Board’s decision and order, and the NLRB cross-petitioned for enforcement. DISCUSSION Courts afford the NLRB broad latitude to develop and implement the procedures for certification elections. NLRB v. Parsons School of Design, 793 F.2d 503, 507 (2d Cir.1986) (citation omitted); NLRB v. Lorimar Productions, Inc., 771 F.2d 1294, 1300 (9th Cir.1985). The Board’s discretion in this area stems from its function to expedite labor elections and to eliminate meritless appeals designed to delay those elections. Hamilton Test Systems, New York, Inc. v. NLRB, 743 F.2d 136, 140 (2d Cir.1984). Thus, the filing of a request for review of a regional director’s decision does not stay an election authorized by the regional director unless the NLRB orders otherwise. 29 C.F.R. § 102.67(b); see Hamilton, 743 F.2d at 140. In addition, the determination of a unit’s composition need not be made before the election so long as the regional director develops “ ‘a balloting process under which the election c[an] be held, whatever the eventual decision might be on the appropriate unit.’ ” Hamilton, 743 F.2d at 141 (quoting Chairman’s Task Force on the NLRB, Interim Report and Recommendations 12 (1976)). The procedure devised must permit the employees to make an informed choice of a collective bargaining representative. Parsons, 793 F.2d at 507; Hamilton, 743 F.2d at 141 (process must “at the very least” inform “voters of the possible alternatives in the selection of the bargaining unit and [permit] them to vote accordingly”). In Parsons and Hamilton, the election notice from the NLRB to the employees described a bargaining unit different from the one ultimately established and did not alert employees to the possibility of change. Par *56 sons, 793 F.2d at 504-05; Hamilton, 743 F.2d at 138. In the instant case, the Regional Director exercised the prerogative of withholding a determination on the unit placement of the automotive floor sales employees until after the election. This decision required the Director to devise a voting procedure to ensure that future placement of those employees would not adversely effect the conduct of the election, regardless of the eventual unit composition. Parsons, 793 F.2d at 506; Hamilton, 743 F.2d at 140. Under the procedure constructed, the automotive floor employees cast ballots in the elections held for both units within the store. If those ballots could change the result of either unit’s balloting, a proceeding would follow to decide on proper unit placement for those five employees. After placement, those votes would be added to the tally from the workers already included in the unit to decide the outcome. The union’s margin of victory in the backshop election triggered this proceeding, and resulted in the inclusion of the automotive floor employees in the back-shop unit. With the ballots of these employees, the union still prevailed. It becomes our task to determine whether the procedure “informed the voters of the possible alternatives in the selection of the bargaining unit and ... permitted them to vote accordingly.” Hamilton, 743 F.2d at 141; see Parsons, 793 F.2d at 507; Lorimar, 771 F.2d at 1302. In doing so, we first consider whether the employees received sufficient notice to understand the voting procedure in general and the unit’s status in particular. Parsons, 793 F.2d at 506-07; Hamilton, 743 F.2d at 140-41. If the notice failed adequately to inform the voters, we examine certain indicators to determine whether employees would have changed their votes had they been better informed. Parsons, 793 F.2d at 507; Hamilton, 743 F.2d at 141. Several weeks prior to election day, the Regional Director sent to Sears for posting in the workplace an election notice explaining the nature of the elections and the disputed unit placement of the automotive floor sales employees. Sears incorrectly posted the notice, leaving hidden the information describing the automotive floor employees’ position. The election monitor rectified the error shortly before balloting commenced by exposing the attachment and orally notifying voters of the existence of the newly exposed portion of the notice. In addition, the Sears employee relations representative explained the procedure to employees at a meeting several days prior to the election. While it is clear that the NLRB may not “inform employees that they are voting for representation in [one] unit and later [] consider th[at] ballot as a vote for representation in [another] unit,” Hamilton, 743 F.2d at 140, that did not occur here. The corrected notification here stands in contrast to the notification given in the elections overturned in Hamilton, Parsons, and Lorimar, where the information distributed turned out to be inaccurate. See Hamilton, 743 F.2d at 138 (election notice informed employees of unit composition but NLRB changed composition after election); Parsons, 793 F.2d at 505 (same); Lorimar, 771 F.2d at 1297 (same). Even in the face of Sears’ careless posting, once the notice was corrected, employees had access to correct information and voiced no objections or confusion before or after balloting. The record established by the NLRB indicates that Sears had already informed voters, at a meeting with the employee relations representative, of the withheld unit placement decision, and that many employees learned of the special position of the automotive floor employees before voting commenced. No inquiry indicating confusion regarding the matter was made. “It seems reasonable to assume that if [some employees] were confused about the unit size, and if that confusion mattered to their votes, some [employees] would have made an inquiry.” See Nightingale Oil Co. v. NLRB, 905 F.2d 528, 533 (1st Cir.1990). Finally, the NLRB rejected Sears’ reliance on its own error in posting the notice as grounds for setting aside the election. The middle panel of the notice was marked to indicate the presence of the at *57 tachment, and the burden of proper posting generally falls on the employer. Contractors Redi-Mix, Inc., 250 N.L.R.B. 121, 121-22 (1980). Any confusion on the part of Sears should have resulted in contact with the Regional Director. For Sears to ignore the situation and now complain about it is impermissible. Contrary to Sears’ suggestion, it is not at all apparent that the voters in this case were misled about the size and composition of the bargaining unit. Nonetheless, we embark on a review of the factors identified in Parsons for determining the validity of election procedures implemented by a regional director in cases of this kind. The factors are: (1) the difference in size between the bargaining unit proposed to the employees before the election and the final size of the unit; (2) the character and scope of the pre- and post-election unit; and (3) the closeness of the election results. Parsons, 793 F.2d at 507 (citing Hamilton, 743 F.2d at 140-41; Lorimar, 771 F.2d at 1302). These factors do not comprise an exclusive list of factors to be considered, but rather must be applied in the context of the circumstances surrounding the election. See Nightingale, 905 F.2d at 535-37; Parsons, 793 F.2d at 507-08; Hamilton, 743 F.2d at 141. The first factor requires examination of the post-election numerical modification of the bargaining unit. Parsons, 793 F.2d at 507 (citing Hamilton, 743 F.2d at 140-41; Lorimar, 771 F.2d at 1302). Generally, when the ultimate size of the bargaining unit is significantly smaller than the unit proposed to the employees before the election, an invalid electoral process is indicated. Lorimar, 771 F.2d at 1302 (invalidating election where unit size reduced by more than one-third); Hamilton, 743 F.2d at 140 (invalidating election where unit size reduced by more than one-half). The apparent rationale is that a smaller bargaining unit may be less attractive to potential union members because of reduced bargaining power. Thus, certain employees voting in favor of what they thought would be a larger, more powerful unit, would have voted against union certification i had they known that the unit would lack sufficient size and strength to justify representation. Hamilton, 743 F.2d at 141; see Lorimar, 771 F.2d at 1302 (citing Hamilton, 743 F.2d at 141). Here, unlike the situations in Hamilton or Lorimar, the unit size grew by approximately ten percent with the inclusion of the automotive floor employees. This increase obviates the fear that some workers voting for the union would have changed their ballots to oppose the union had they been informed of the smaller size of the eventual unit. Indeed, the ultimate unit composition might have led more employees to vote in favor of the union for the perceived benefits from greater size. We are therefore convinced that this consideration supports the Regional Director and fails to indicate that some votes would have changed to union opposition. Even if we accepted Sears’ proposition that it is irrelevant whether the unit increases or decreases in size after the election, a ten percent change does not justify invalidating the process. See Parsons, 793 F.2d at 507. The court in Parsons noted that a ten percent decrease lacks the numerical significance of Hamilton and Lorimar and requires an examination of the other two factors in order to invalidate the election. Id. The next indicator focuses on the change in character and scope of the unit before and after modification. Parsons, 793 F.2d at 507 (citing Hamilton, 743 F.2d at 140-41; Lorimar, 771 F.2d at 1302). This factor requires consideration of the similarity or dissimilarity of job classifications between the original and final units and the chance that the ultimate unit split an otherwise unified workforce. Id.; see Hamilton, 743 F.2d at 141; Lorimar, 771 F.2d at 1302. Job classification differences include variations in both salary and . type of work. Hamilton, 743 F.2d at 141. Also important is the possibility that unit composition would lead to interpersonal friction within the unit, and that such a result would have been considered by voters. Parsons, 793 F.2d at 507; Hamilton, 743 F.2d at 141. If the new composition of the unit excludes workers performing under *58 conditions similar to those already included in the unit, employees are more likely to have voted differently because they may feel that their individual interests cannot be best represented by a group consisting of diverse, potentially adverse interests. Id.; Parsons, 793 F.2d at 507. In this case, Sears initially sought to have the automotive floor sales employees included with the backshop staff in the bargaining unit. This is strong evidence that no great change in scope or character of the unit occurred with the late addition of the automotive sales employees. Furthermore, unlike in Parsons, the floor employees perform the same functions as certain backshop workers, albeit in a different part of the store. Parsons, 793 F.2d at 507 (noting that group excluded from final unit differed from those remaining in final unit). For example, the floor employees provide the same advice to customers, are paid on the same commission basis as the backshop service advisors, and sell the same automotive products. The same person supervises the floor employees and all non-mechanic backshop employees. Thus, the scope of the unit is unchanged. Finally, the introduction of the automotive floor sales employees will not cause the disruption of a unified workforce as occurred in Hamilton, 743 F.2d at 141 (noting that some employees may have voted for union with expectation that workforce would remain unified because originally proposed bargaining unit comprised all employees at location), and Parsons, 793 F.2d at 507 (same). In fact, the converse is true. By including the automotive floor staff, the final unit represented a unified automotive workforce and was not likely to create interpersonal frictions or interest-group power struggles. This examination convinces us that had the employees been better informed about the unit’s ultimate composition, the negligible alterations in character and scope would not have led to changed votes warranting a finding of an invalid election. The final Parsons indicator is the margin of victory. Parsons, 793 F.2d at 507 (citing Hamilton, 743 F.2d at 141; Lorimar, 771 F.2d at 1302). A narrow victory heightens the need to scrutinize the election process to ensure that votes would not change with a more fully informed electorate. Id. This factor must be observed in combination with the two addressed above to see if the concerns uncovered in examining the other factors would have led to a vote change sufficient to alter the election outcome. Id. at 508; see Hamilton, 743 F.2d at 141. The union won the Sears backshop unit election by twenty-five votes to twenty-three. As in Hamilton, “[a] change of one vote from union to non-union would have altered the outcome of the election.” Hamilton, 743 F.2d at 141. But unlike Hamilton, an analysis of the other two factors does not reveal concerns within the workforce justifying the assumption that any votes would have changed. Standing alone, the small margin does little to prove that, had the backshop staff definitely known the true placement of the floor employees, the result would have been different. It seems unlikely, contrary to Sears’ suggestion, that anti-union people simply voted for the union because they believed exclusion of the floor employees made approval of the union a foregone conclusion. The other reasons offered by Sears as to why employees might have changed votes rely on the differences between the floor employees and the backshop staff. In light of the overlapping functions of various backshop personnel and the floor staff, it is difficult to see how this would have made a difference in the election. As a consequence, we find that consideration of the factors dictated by our prior holdings leads to the conclusion that the election process chosen by the Regional Director was not marred by the post-election determination of the automotive floor sales employees’ unit placement. Sears also contends that the breakdown of votes within the eventual unit demonstrates that better notice would have resulted in changed votes. The automotive sales employees were fully informed of the voting procedure throughout the election. *59 They voted against the union four to one. Meanwhile, the majority of the employees in the backshop voted in favor of the union despite the defective notice. From these facts, Sears concludes that better informed voters rejected the union and thus, had the remaining voters been apprised of the situation, they too would have rejected the union. This argument fails in several ways. First, Sears failed to prove that any specific votes would have changed. Without such evidence, its argument rests on pure speculation. Second, we have already demonstrated the absence of reasons why the addition of the automotive floor sales employees would have altered any backshop staff votes. Sears’ simply argues that the automotive floor employees did not favor the union, regardless of the unit composition. This says nothing about the voting preference of the backshop staff. “ ‘The test is not whether optimum practices were followed, but whether on all the facts the manner in which the election was held raises a reasonable doubt as to its validity.’ ” Nightingale, 905 F.2d at 531 (quoting NLRB v. ARA Services, Inc., 717 F.2d 57, 68 (3d Cir.1983)). When all the surrounding facts and circumstances are considered, we cannot say that the manner in which the election was held raises reasonable doubts as to its validity. Sears did not carry its heavy burden in challenging the election, and the NLRB did not abuse its discretion in finding the election process acceptable. CONCLUSION The NLRB order directing Sears Roebuck & Co. to bargain with the union is enforced.