I dissent. I do not object to the majority’s announced test for review of campaign threats—“whether the statements would reasonably tend to interfere with or restrain employees in the exercise of their rights guaranteed by the Act.” (Jack Brothers and McBurney, Inc. (1978) 4 A.L.R.B. No. 18 at p. 3.) However, in applying the test, the majority violates a fundamental principle of judicial review in substituting its own judgment for that of the board in order to justify the attack on findings supported by substantial evidence. The majority freely acknowledges that “[t]he findings of the board with respect to questions of fact if supported by substantial evidence on the record as a whole shall ... be conclusive.” (Lab. Code, § 1160.8; Tex-Cal Land Management, Inc. v. Agricultural Labor Relations Bd. (1979) 24 Cal.3d 335, 345 [156 Cal.Rptr. 1, 595 P.2d 579].) There is substantial evidence in the record to support the conclusion that employees would consider the statements to be campaign propaganda which the union could not effectuate, and the board did not abuse its discretion in concluding in effect that the statements did not reasonably tend to interfere with the employees’ free exercise of their rights under the Agricultural Labor Relations Act.
The test utilized by the board to review the statements has already been approved by this court in J.R. Norton Co. v. Agricultural Labor Relations Bd. (1979) 26 Cal.3d 1, 22 [160 Cal.Rptr. 710, 603 P.2d 1306], That test focuses on whether there is evidence “suggesting the activity had a potential *58for interfering with the employee’s free choice.” The test stated by the majority for cases involving campaign threats—“whether the statements would reasonably tend to interfere with or restrain employees in the exercise of their rights guaranteed by the Act”—is not substantively different from the test already approved by this court for measuring other election misconduct and applied by the board in the instant case.
The board said “the statements indicate that the employee will be replaced simply for voting against the union. The implication is that the union knows how the workers vote and that the union somehow has power over job tenure or discharge, regardless of whether it wins the election. [¶] The NLRB has held that organizer statements, which are not accompanied by some indication of the union’s ability to carry out the suggested threat, are not likely to affect the employees’ free choice of representative. [Citations.] Given the circumstances herein and the character of the statements made, we find that such statements would be viewed by the employees as campaign propaganda which the Union could not effectuate. Further, the record shows that only a small number of employees heard the statements. Therefore, we conclude that such statements did not influence or affect the employees in their choice of a bargaining representative. ”
The obvious import of the board’s opinion is that it would be unreasonable to conclude that in a secret election the union would know how people vote and be able to influence whether particular voters retained their jobs even if the union lost the election. Finding such conclusion is not reasonable, the board could properly refuse to accept it. Despite the board’s obvious disbelief that the voters could reasonably be coerced by statements implying that “the employee will be replaced simply for voting against the union,” (maj. opn. at p. 51), the majority amazingly declares that “the Board recognized that the reasonable implication from such a consequence was that the union would both know of each worker’s vote and would exercise some control over job tenure.” The majority having fundamentally misread the board opinion (see maj. opn. at pp. 51, 53, 54), it is not surprising that it concludes that the board erred in dismissing the employer’s objections.
The determination whether statements reasonably tend to produce a coercive effect on the free exercise of employees’ rights is a judgment call better left to the expertise of the board. It involves a decision that requires the drawing of inferences based on the peculiar facts of each case placed in the context of the accumulated experience of the board. (Abatti Farms, Inc. v. Agricultural Labor Relations Bd. (1980) 107 Cal.App.3d 317, 327 [165 Cal.Rptr. 887].) “[W]hether certain words do contain a threat of reprisal *59and would reasonably tend to restrain or interfere with an employee’s exercise of a protected right is a matter that calls for some detailed understanding of the agricultural labor scene, the peculiar relationship between a field laborer and supervisor with power to hire and fire the non-English speaking agricultural migrant laborers, as well as the nature of the symbols used to communicate. These questions presented to the Board were of such nature as to call for expertise. Presumptively, the Board is equipped or informed by experience to deal with this specialized field of knowledge; therefore their findings carry the authority of an experience which the courts do not possess and therefore must respect. (Universal Camera Corp. v. Labor Board (1951) 340 U.S. 474, 488 [95 L.Ed. 456, 467, 71 S.Ct. 456].)” (Abatti Farms, Inc. v. Agricultural Labor Relations Bd., supra, 107 Cal.App.3d 317, 340 (conc. opn. of Staniforth, J.).) The deference to be given board inferences and conclusions is equally applicable to situations involving review of alleged misconduct by union organizers.
Although this court might disagree with the board as to the inferences to be drawn from the evidence in this case, there is substantial evidence in the record to support the inferences drawn by the board; that being the case, we should not disturb its judgment. “[I]f there is evidence to support each of two conflicting views, the findings of the Board must be allowed to stand despite the fact that we might have reached the opposite conclusion on our own. [Citations.]” (N. L. R. B. v. Pacific Grinding Wheel Co., Inc. (9th Cir. 1978) 572 F.2d 1343, 1347.)
Because it is unlikely that there would ever be direct evidence that particular conduct influenced the employees in their choice of a bargaining representative, the board must necessarily draw inferences from the character and circumstances of the objectionable conduct. In analyzing the conduct in a particular case, the board must be permitted to calculate the effect of misconduct based on an objective test developed from the board’s cumulative experience with union representation contests. Otherwise, objection hearings would degenerate into extended parades of witnesses for the union and the employer, each testifying whether he or she was or was not intimidated by the conduct. Moreover, such an inquiry could result in violation of the right of employees to secrecy of the ballot.
The board concluded that the statements could not be said to reasonably tend to interfere with the employees in the exercise of their rights, because they were not accompanied by some indication of the union’s ability to carry out the threat. Because the statements here were not tied to any conduct by *60the employees which would verify their responses to the statements, the statements are clearly less coercive than those requiring some overt act such as signing union authorization cards. Moreover, the board correctly concluded that such statements would not reasonably be coercive when they were not conditioned on the union winning the election, because the union was obviously powerless to effectuate threats if it lost the election.
The majority rejects the board’s analysis of the statements in favor of its own conclusions based on its misreading of the board opinion. The majority’s attempt to distinguish the cases relied on by the board does nothing to undercut its conclusion that the statements would be viewed as campaign propaganda which the union could not effectuate.
Nor are the board’s conclusions undercut by the majority’s invocation of the supposed National Labor Relations Board (NLRB) “rule” that “ ‘statements made during an election can reasonably be expected to have been discussed, repeated, or disseminated among the employees . . . .’” (Maj. opn. at pp. 52-53.) The majority portrays this statement as if it is a general NLRB rule, but examination of the cases cited to support the proposition suggests that the “rule” has been applied in quite limited circumstances dissimilar to those here.
The threat made in United Broadcasting Company of New York (1980) 248 N.L.R.B. 403 was made to one employee in a bargaining unit of six employees of whom four cast ballots. In Standard Knitting Mills, Inc. (1968) 172 N.L.R.B. 1122, although the statements were made to 4 employees out of 3,000, the union lost the election by only 17 to 21 votes. “In essence, therefore, the impact of the misconduct is the same as if the entire unit consisted of, at most, 21 employees.” (172 N.L.R.B. at p. 1123.) The threats made in Sav-On-Drugs (1977) 227 N.L.R.B. 1638 were made to 6 employees in a unit of 38. None of these cases represent “applicable NLRB precedents” warranting the majority criticism of the board. (Maj. opn. at p. 52.) In the instant case there were 420 eligible voters, the threats were only made to a few, and the election was not close.
Finally, the majority improperly elevates its application of the supposed NLRB presumption of dissemination to the status of evidence. By this device, the majority concludes that the testimony regarding the effects of threats on witnesses and coworkers supports a finding that the statements created an atmosphere of fear and coercion. The majority reaches this con*61elusion despite the testimony of the employer’s witnesses that they knew of no workers who failed to vote because of the alleged intimidation.
As the majority acknowledges, “[hjearsay evidence may be used for the purpose of supplementing or explaining other evidence, but shall not be sufficient in itself to support a finding unless it would be admissible in civil actions.” (Cal. Admin. Code, tit. 8, § 20370, subd. (c), italics added.) The majority also concedes that the testimony that other workers were afraid would not, standing alone, be sufficient to invalidate the election. (Maj. opn. at p. 55.) But, the majority couples the hearsay evidence with the NLRB presumption to create its own finding that the statements induced fear on the part of the workers (maj. opn. at p. 56).
The fatal flaw in the majority’s reasoning is that a presumption, even properly applied, is not evidence. (Evid. Code, § 600, subd. (a); People v. Hedrick (1980) 105 Cal.App.3d 166, 171 [164 Cal.Rptr. 169].) Thus, the presumption does not substitute for the “other evidence” necessary to support the majority’s “finding.” The majority has merely bootstrapped hearsay evidence and an inapplicable presumption to support its judgment, not otherwise supported by the record, that the statements “created an impermissible atmosphere of fear.” (Maj. opn. at p. 56.)
The majority’s purpose in this case seems clear. It is determined that any threats against employees necessarily are coercive and cause to invalidate the election, despite the context in which they were made, the number of employees who heard the threats, the patent inability of the union to effectuate the threats, or the large vote in favor of the union. Thus the majority rejects the reasonable inferences of the board in favor of its own conclusion that threats necessarily interfere with or restrain employees in the exercise of their rights guaranteed by the act.
Of course, such conduct as occurred in this case is not to be condoned. However, I do not foresee upholding the board’s decision in this case as an invitation to similar conduct in the future by competing unions and employers. The board is best equipped to evaluate such conduct based on its experience in the field of agricultural labor relations, and to refuse to certify an election when “misconduct affecting the results of the election occurred.” (Lab. Code, § 1156.3, subd. (c). When the record supports the board’s findings and inferences, as in this case, its decision should not be upset merely because this court has a different view of the “proper” result.
I conclude that there is substantial evidence to support the findings and inferences of the board that the statements did not influence or affect the *62employees in their choice of a bargaining representative. Thus, I would affirm the decision of the Agricultural Labor Relations Board.
Reynoso, J., concurred.
The petitions of respondent and real party in interest for a rehearing were denied January 19, 1984. Bird, C. J., did not participate therein. Broussard, J., and Reynoso, J., were of the opinion that the petitions should be granted.