HIGH MIGHTY FARMS v. UNITED FARM WORKERS OF AMERICA AFL CIO (1983)

OPINION

Recognizing the need to provide stability and to reduce tension in the California farming industry, the Legislature enacted the Agricultural Labor Relations Act (ALRA) in 1975.   One of the avowed purposes of the act was to provide for maximum participation by farm workers in selecting their representative for collective bargaining.   Realizing a farmer's labor force is comprised of both regular and seasonal employees and also varies with the season, the Legislature provided union elections could be held only when the total employment level was at least 50 percent of peak.   This appeal challenges the Agricultural Labor Relations Board's (ALRB or Board) certification of a bargaining representative.   Petitioner asserts the ALRB's multiple approaches to peak calculation applied in ad hoc fashion without predetermined standards inject so much uncertainty in the election process that they are inconsistent with the ALRA's purpose of bringing peace and stability to the fields ¹ and must be deemed arbitrary.

BACKGROUND AND PROCEDURAL HISTORY

Petitioner, High and Mighty Farms is a California corporation growing and harvesting lettuce and other row crops in California and Arizona.   High and Mighty employs regular and seasonal employees hired directly and paid from its general payroll.   Additional seasonal employees are hired through a labor contractor who is paid by High and Mighty and who in turn pays these workers.   This case deals with the attempt by the United Farm Workers of America, AFL–CIO (UFW) to organize these High and Mighty employees.

Chapter 5 of the ALRA (Lab.Code, § 1156 et seq.) ² includes the procedures necessary to the election and certification of an agricultural workers' union.   The process begins by the filing of “Petition for Certification” which must set forth a sufficient showing of interest and support by a majority of currently employed workers.  (§ 1156.3, subd. (a);  Cal.Admin.Code, tit. 8, § 20300.) ³  Employees manifesting their support for an election submit to the party filing the petition an authorization card.   The ALRB, through a regional director, then calculates the number of employees working each day of the payroll period, adds them together, then divides by the total number of days in the period.   A simple majority of authorization cards, when compared to this average is sufficient to meet this employees' support requirement.  (See Cal.Admin.Code, tit. 8, § 20300, subd. (j)(1).) ⁴

The petition must also allege, inter alia:  “That the number of agricultural employees currently employed by the employer named in the petition, as determined from his payroll immediately preceding the filing of the petition, is not less than 50 percent of his peak agricultural employment for the current calendar year.”  (§ 1156.3, subd. (a)(1).)   This 50 percent requirement is the focal point of the present controversy.   Unlike the employee support requirement there is no comparable administrative code section outlining how this calculation is to be made.   However, the Legislature, in section 1156.4 has made one other reference to this peak requirement:

“Recognizing that agriculture is a seasonal occupation for a majority of agricultural employees, and wishing to provide the fullest scope for employees' enjoyment of the rights included in this part, the [Board] shall not consider a representation petition or a petition to decertify as timely filed unless the employer's payroll reflects 50 percent of peak agricultural employment for such employer for the current calendar year for the payroll period immediately preceding the filing of the petition.

“In this connection, the peak agricultural employment for the prior season shall alone not be a basis for such determination, but rather the [Board] shall estimate peak employment on the basis of acreage and crop statistics which shall be applied uniformly throughout the State of California and upon all other relevant data.”

On November 17, 1975, the UFW filed its petition for certification alleging all the requirements for a union election had been satisfied.   A November 24, 1975, election resulted in the UFW acquiring a majority of votes, sufficient to elect the UFW as the sole bargaining agent for High and Mighty employees.⁵

High and Mighty filed objections to this election under the applicable provisions of the ALRA and the Administrative Code.   Among the objections was a charge that the Board agent abused his discretion by failing to dismiss the certification petition because employees currently employed in the last payroll period, prior to the filing of the certification petition, did not represent 50 percent of the employer's peak employment for the current year.   Because we have concluded this objection is well taken and a new election must be held, we need not deal with the other objections.

An April 14, 1977, hearing before an Investigative Hearing Examiner (IHE) resulted from the timely filing of these objections.   While several documents were received into evidence and a written stipulation of facts as well as oral stipulations were made, no witnesses testified.   In a 22-page opinion (75–RC–10–I), the IHE recommended High and Mighty's objections be dismissed and that the UFW be certified as the exclusive bargaining representative.   In High and Mighty Farms, Inc. (1975) 3 ALRB No. 88, the ALRB followed the recommendation.

 Because an order certifying a union election is not directly appealable (Nishikawa Farms, Inc. v. Mahony (1977) 66 Cal.App.3d 781, 787, 136 Cal.Rptr. 233;  Perry Farms, Inc. v. Agricultural Labor Relations Bd. (1978) 86 Cal.App.3d 448, 470, 150 Cal.Rptr. 495), to obtain review High and Mighty purposefully refused to bargain with the UFW, resulting in the filing of an unfair labor practice charge.

In High and Mighty Farms (1978) 4 ALRB No. 51, the ALRB found High and Mighty guilty of unfair labor practices, awarding make-whole relief and requiring High and Mighty to bargain with the UFW.   After a Petition for Writ of Review was denied by this court, High and Mighty petitioned the Supreme Court.   That court ordered us to reconsider in light of J.R. Norton Co. v. ALRB (1979) 26 Cal.3d 1, 160 Cal.Rptr. 710, 603 P.2d 1306.   We remanded this case to the ALRB who in High and Mighty Farms (1980) 6 ALRB No. 31 reversed its make-whole award but affirmed its certification of the UFW.   This case is now properly before us for the review of the objections made concerning the election.  (J.R. Norton Co. v. ALRB, supra, 26 Cal.3d at p. 18, 160 Cal.Rptr. 710, 603 P.2d 1306.)

PEAK CALCULATIONS

A review of the evidence before the IHE and the ALRB reveals the following:

High and Mighty's peak employment period for 1975, as evidenced by its payroll records was July 2 through and including July 8.   In this payroll period all workers were hired directly.

b⁶ ,⁷

High and Mighty's regular payroll period ending immediately before the filing of the Petition for Certification encompassed November 5 through and including November 11.   In addition, High and Mighty employed contracted laborers who were paid by the contractor whose pay period went from November 6 through and including November 12.

b⁸ ,⁹

In ascertaining the petitioner's peak employment, the IHE used an averaging method.   The number of persons working each day during the peak week was totaled and then divided by seven, the number of days in the pay period.   The IHE concluded the Arizona workers should not be included in the peak calculations and, excluding them, found the total number of workers for the peak week was 685.   Then dividing this number by 7, the IHE found High and Mighty employed on the average 98 ¹⁰ persons per day during the peak period.  (Had the Arizona workers been included in this calculation, a total of 937 results from adding the daily total.   Dividing 937 by 7, the IHE would have found the average to be 134.)

The IHE then calculated the number of workers employed during the payroll period immediately preceding the filing of the petition.   This figure must be at least 50 percent of the peak week figure in order to certify the elected union.  (§ 1156.3, subd. (a)(1).)   As the chart above shows, during that payroll period High and Mighty employed a group of “regular” workers, paid directly through its own payroll, and “contract” workers, who were paid by the contractor.   High and Mighty's pay period ran from Wednesday to Tuesday while this particular contractor's pay period ran from Thursday to Wednesday.   Adding the totals of the regular employees for this period resulted in a figure of 83.   83 divided by 7 gave an average of 12.   The IHE then concluded because the contractor had a different pay period when compared to the employer, a separate calculation of the contracted workers was necessary.   Hence, ignoring the days of High and Mighty's pay period for the time being, the IHE found the last full pay period of the contractor and totaled the number of persons who worked each day of this period.   This total was 233.   Even though this pay period encompassed 7 days, the IHE divided 233 by 3 because no contract workers were employed during the first 4 days of this period.   This total was 78.   Adding these two averages together resulted in 98.   The method used by the IHE resulted in the conclusion that 98 persons worked on an average day during the peak period and, excluding four contract labor days, an average of 98 persons worked during the pre-election week.   In other words, employment during the week preceding the filing of the petition for certification was 100 percent of peak.   Petitioner suggests the ALRB has, in effect, worked backward, adopting a policy of using whatever formula is necessary to achieve a result favorable to the union.   This contention is engendered by the ALRB's practice of implementing different formulae for calculating peak to deal with different circumstances surrounding a particular employer.

PAST ALRB DECISIONS ON PEAK

As previously indicated, there are two statutory provisions discussing the 50 percent peak requirement (§§ 1156.3, subd. (a)(1) & 1156.4).   The ALRB has employed several methods in determining peak within the statutory requirements.   The first method was simply a process of counting and comparing the number of workers during the pre-election period and peak period.   The director attempted to use this approach in Mario Saikhon, Inc. (1976) 2 ALRB No. 2.   However, the ALRB concluded a large daily turnover in the work force justified a different approach.   Consequently, an averaging formula was devised:

“[W]here an employer employs the same 100 individuals each day for the five-day period, there would be 100 employee days worked each day and the average number of employee days worked would be 100.   Where an employee has total turnover and employs 100 different employees each day, there would be also 100 employee days worked each day and the average number of employee days worked on all the days of the payroll period would be 100.   Thus, the approach we adopt of taking the average of the number of employee days worked on all the days of the payroll period avoids the pitfalls of the ‘employee count’ method by yielding a consistent measure despite employee turnover.”

In Ranch No. 1, Inc. (1976) 2 ALRB No. 37, decided 47 days after Saikhon, the ALRB found:  “[T]hat although 107 employees worked during the payroll period ending September 21, 1975, the average number of employee days worked that week was only 59.5 4 because many employees worked for only one or two days.   For the payroll period ending August 10, 1975, during which 241 employees worked, the average number of employee days worked was 191.”  (Fns. omitted.)

“4 The average number of employee days for each of the two payroll periods was calculated by determining the actual number of employees who worked on each day of the week, then adding the totals for the six days of Monday through Saturday and dividing by six.   Sunday was not added in for either period because only a few employees worked on each Sunday so that the addition of Sunday and division by seven would yield an average number of employee days which is not representative of the average of the other six days․”

In Ranch No. 1, the ALRB rejected the UFW's request that peak be estimated in accord with section 1156.4.  “We think the clear import of that provision is that the Board is required to take into account crop and acreage statistics only when it is alleged that peak will occur at some future point in the calendar year.   In such circumstances, reliance on employment records for the prior season might well be inadequate to project peak for the current year.   However, where, as here, it is contended that peak employment has already occurred within the current calendar year, a comparison between employment figures in the two relevant payrolls will fully reveal whether the petition for certification was timely filed.   No supplemental data concerning crop or acreage statistics is necessary to make the purely mathematical computation of whether the payroll for the period immediately preceding the filing of the petition was 50 percent of the payroll in the earlier period claimed to constitute peak.”

In Luis A. Scattini & Sons (1976) 2 ALRB No. 43, the ALRB had before it two different groups of employees, regular and contract.   The regular employees were paid every two weeks and the contract employees were paid daily.   After discussing the approach taken in Saikhon and Ranch No. 1, another approach of calculation of peak was suggested.  “An alternative approach is to compute the average number of employee days worked separately for the two classes of employees.   For the regular workers, that figure would be computed over the relevant two-week payroll periods, since the regular workers are paid on a two-week basis.   For the labor contractor employees, paid on a daily basis, we might proceed by analogy to Section 23055 of our regulations (8 Cal.Admin.Code, § 20355), which provides that where an employer's payroll is for fewer than five working days, the relevant payroll period will be presumed to be at least five days long.   Using this approach for the labor contractor employees, we would compute the average number of employee days worked over a period of five working days.   The ‘average’ figures for the two types of employees would then be added together to reach an overall figure for this period.

“Under this alternative approach, during the period alleged to constitute peak, we would use statistics from the five consecutive days with the highest number of labor contractor employees.   For the comparative period preceding the filing of the petition, two methods of computation are possible:  (1) use the five consecutive days of highest labor contractor employment within the two-week payroll period preceding the filing of the petition, or (2) follow the literal wording of section 20355, and use employment figures from the five working days immediately prior to the filing of the petition, regardless of whether those days fall within the two-week payroll period preceding the petition's filing.   Whichever period is used, the average number of employee days worked by regular employees would then be added to the average number of employee days worked by labor contractor employees.”

In Valdora Produce Co. (1977) 3 ALRB No. 8, the ALRB determined peak utilizing the straight count method.   The record indicated the employer employed during the peak period 329 workers and during the pre-election period 166 workers.   The straight count method is different from the employee count.   The former merely involves adding the number of workers employed each day while the latter involves calculating the number of different persons employed during the respective period.   Without any further analysis, the Board determined 166 was more than 50 percent of 329 and dismissed the employer's objection.

In Kawano Farms, Inc. (1977) 3 ALRB No. 25, the information provided by the employer indicated 1975 peak employment consisted of 930 employees and the pre-petition period consisted of 649 employees.   Because the pre-petition was filed on September 8, 1975, before the end of the calendar year, the employer contended the Board agent could not determine at least 50 percent of peak employment for the current calendar year had been reached.   The Board rejected this contention utilizing the following language:  “In Ranch No. 1, Inc., 2 ALRB No. 37 (1976), we disapproved the contention that Section 1156.4 prevents Board agents from relying solely on the employer's employment records in determining whether peak had already been reached for the current calendar year.

“The regional director was free to rely on the two relevant payrolls supplied him.   Given the 649 employees, the employer was well at peak.”

In Donley Farms, Inc. (1978) 4 ALRB No. 66, the record indicated Donley employed 227 different employees during the peak period and 121 different employees during the pre-petition period.   Accordingly, under the employee count method, Donley would have been at 50 percent peak immediately preceding the filing of the petition.   Applying the Saikhon approach would have resulted in a calculation of an average of 73 employee days for the pre-petition period and average of 181 employee days for the peak period, far less than 50 percent of peak.   The UFW contended the employee count method should be utilized and the employer contended the Saikhon method should be utilized.   The IHE rejected the employer's contention and accepted the UFW's contention, utilizing this language:  “The legislature was attempting to insure that the actual voters in the election were of a sufficient number to be representative of all the workers in expressing their desire to have an election, regardless of outcome.   Since every agricultural worker who was employed during the pre-petition period is given the right to vote, I find no support within the Act for a proposition by which the value of a worker's interest in the election is dependent upon the number of days he or she works.   This is the inevitable result of averaging the employees in the pre-petition period.   The assumption is that if 6 different workers hold the same job position for a 7 day pay period because of high turnover in that position, they each are less representative of the wishes of the whole workforce during that time than one worker who holds a job position for an entire 7 days.   I find this result incompatible with the Act's stated purpose of ‘guaranteeing justice for all agricultural workers,’ and the legislative scheme under which each is given an equal vote, regardless of number of days worked.”  (Fn. omitted.)

The ALRB considered and affirmed the rulings, findings and conclusions of the IHE, including the refusal to apply the averaging formula:  “The Employer contends that the IHE should have applied the Saikhon method in determining peak.   This method compares the average number of employees working each day during the two relevant payroll periods.   Use of the Saikhon method is unwarranted in the instant case as a conventional count of the number of employees in each of the payroll periods establishes that the Employer was at 53.3 percent of peak during the pre-petition period․”  (Fns. omitted.)

In Bonita Packing Co., Inc. (1978) 4 ALRB No. 96, the pre-petition was filed in September 1975 and the beginning of peak period of employment was designated first as March 8, 1975, and then March 16, 1975.   The pre-petition period showed 58 employees and the peak periods 117 and 119 employees, respectively.   The ALRB, after discussing Labor Code section 1156.4 and the other previous decisions, concluded 50 percent of peak had been reached, utilizing the following language:  “In this particular case, 58 employees were eligible to vote in the election.   Prior to the election herein, the Employer selected the week of March 8, 1975, as representative of its peak employment, and stipulated at the first hearing in this matter that 117 employees worked during that period.   At the second hearing, it designated the week of March 16–23 as representative of its peak employment, and introduced payroll records which show that 119 employees worked during that period.   The record in this case does not reveal why the Employer initially selected the earlier week in March as representative of its peak employment.   It appears, however, that the Employer experienced its peak employment for 1975 over a period of time which was longer than one payroll period.   In this context, the total number of individuals working during a single one of those periods must be taken as an approximate measure of peak employment.   On this record, the fact that the total number of eligible voters fell short of being over 50 percent of 119 by a margin of two employees, does not indicate that this election was unrepresentative and is not adequate reason to refuse to certify its results․”

In Wine World, Inc., dba Beringer Vineyards (1979) 5 ALRB No. 41, the ALRB held that under Labor Code section 1156.4, it was required to estimate peak in situations where the election occurs after the employer has experienced peak as well as in situations where the employer contends peak will be reached after the election in the current calendar year.   The Board further indicated that the difficulty in determining peak allowed them to estimate peak and Ranch No. 1 which required a purely mathematical comparison in past peak employment cases could no longer be followed.   The Board, in setting aside the election, discussed the reasonable margin of error theory adopted in Bonita:  “In Bonita, the 58 eligible employees were 50 percent of 116.   The Regional Director's estimated peak was 119.   The figure of 116 resulted in a margin of error of 2.5 percent in the estimate of 119.   In view of our comparison of a set figure (number of employees eligible to vote) with an estimate (the calendar-year peak figure) and the inherent difficulties involved in making the estimate, we conclude that such a margin of error was reasonable.   In this case, the 60 eligible employees were 50 percent of 120.   The Regional Director's estimate was 129.   The margin of error here is approximately 7 percent.   We find that margin of error unreasonable.   Accordingly, the election will be set aside and the petition for certification will be dismissed.”  (Fn. omitted.)

ISSUES

We are not called upon to pass judgment on the historical development of the Board's various formulae nor to analyze the reasons behind their development.   The cases have been reviewed simply to provide an historical perspective in which to determine whether the formula used in this case comports with the statutory requirement of 50 percent of peak and the statutory purpose embodied in that requirement of providing for the maximum participation of workers in the election process.   In making this analysis, two essential questions must be answered.   First, should the Arizona workers be included in the figures to determine peak?   Second, did the Board correctly determine High and Mighty was at least at 50 percent of peak employment at the time the election was held?

THE ARIZONA WORKERS

High and Mighty raises crops in California and Arizona.   During the peak week of July 2, honeydew harvesting was going on in the Arizona field.   In a joint stipulation between High and Mighty and the UFW, the parties agreed:

1. These Arizona workers are paid by High and Mighty through the Blythe, California, office.   The company does not have an Arizona office.

2. The management and supervisors of these workers are the same individuals that supervise the melon harvest in California during this same period.

3. High and Mighty is a California corporation.

4. The employees harvesting honeydews are paid the same wages and on the same basis as are all other High and Mighty harvesting employees.

5. These employees also work in the melon harvest in California.

6. The company's melon fields in Arizona are 25–30 miles from the California border.

7. These employees commute to work in their own vehicles.

8. California State Disability is withheld from these employees' paychecks, even while they are working on melon fields in Arizona.

9. No Arizona state deductions are withheld from these employees' wages.

The IHE refused to include these Arizona workers in his peak calculations on the basis that “[t]he ALRB has no jurisdiction over operations outside the State of California.”   In addressing the method of determining peak used in High and Mighty Farms, Inc., supra, 3 ALRB No. 88, the ALRB avoided this issue by concluding the addition of the Arizona workers would have had no effect on the conclusion that High and Mighty was over 50 percent of peak during the election period.

Were this a jurisdictional issue we would find no problem in concluding California had “sufficient contacts with and interest in the dispute to assure the application of [its] law is neither arbitrary nor unfair.”  (United Farm Workers v. Arizona Agricultural Employment Relations Board (9 Cir.1982) 669 F.2d 1249, 1256.)   The ALRB in Mario Saikhon, Inc. (1978) 4 ALRB No. 72, acknowledged sufficient contacts was the proper standard to be used in determining whether acts occurring outside the state could be governed by the ALRA.   The facts in the joint stipulation amply satisfy the sufficient contacts requirement.

 The very formation of the employer-employee relationship within California gives the state a sufficient jurisdictional basis, commensurate with due process, to regulate the incidents of that relationship, even when it involves the rendition of services within another state.  (See Alaska Packer's Asso. v. Industrial Acci. Com. of Cal. (1935) 294 U.S. 532, 540 [55 S.Ct. 518, 520, 79 L.Ed. 1044, 1048].)   However, jurisdiction over out-of-state workers is not in issue.   The right to vote is in issue.   The statute requires the employer to be at least at 50 percent of peak before an election may be held.   The ALRA contemplates “elections ․ be conducted [only] when a meaningful number of employee workers are employed and will allow employees to freely decide upon their representation.”  (Cadiz v. Agricultural Labor Relations Bd. (1979) 92 Cal.App.3d 365, 155 Cal.Rptr. 213.)

 Obviously any bargaining agreement would affect the rights of the Arizona workers, thereby destroying one of the most important incidents of the employee-employer relationship, i.e., the right of the employee to select a bargaining agent with the concomitant duty of the employer to bargain with that agent.   The failure to consider the Arizona workers in determining peak employment was tantamount to depriving them of the right to vote.   They were deprived of the right to participate in the selection of the bargaining agent.

We hold for purposes of determining the peak period of employment the California workers temporarily working in Arizona form a part of those employees who must be included.

PROPRIETY OF THE BOARD'S ORDER DETERMINING PEAK

In order to conclude that the pre-petition employment was at 50 percent of peak or more in this case the IHE and the Board found it necessary to utilize a combination of several of the approaches discussed in our review of the cases.   The IHE used Saikhon as authority for averaging;  Ranch No. 1 to justify using three as a divisor instead of seven to average the number of contract workers;  and Scattini to calculate the averages of the regular and contract workers separately and add them together.   As demonstrated below, none of these formulae standing alone, nor any of the previously devised formulae, would produce the result necessary to uphold the election:

(a) Saikhon:  Utilizing a seven-day work week as the divisor for 937 workers during peak week results in an average of 134 workers;  averaging 316 pre-election week workers with 7 as the divisor equals 45 workers.   Because 45 workers is less than 50 percent of 134, the election could not be certified.

(b) Ranch No. 1 provided unrepresentative days be dropped from the divisor.   All seven days of the work week during peak period were well represented—the average of 937 is 134.   Given the record of employment for the pre-election period, it is difficult to conclude any particular day was unrepresented.   Nevertheless, even if we assume 2 days were unrepresented, utilizing 5 as a divisor results in an average of 63.   63 is less than 50 percent of 134, the election could not be certified.

(c) Scattini:  Utilizing a seven-day pay period for the divisor, 134 results from averaging 937 workers.   Using separate calculations for the pre-election week reveals 233 contract workers with a resulting 33 average;  83 regular workers with a resulting 12 average.   The combined 45 average is less than 50 percent of 134 and the election could not be certified.

(d) Valdora and Kawano utilized the straight-count method without averaging.   Here, 316 (pre-petition workers) is less than 50 percent of 937 (peak-week workers).   So the election could not be certified.

(e) Donley:  Utilized the employee-count method, i.e., counting the number of different workers employed during peak week and the number of different workers employed during pre-election week for comparative purposes.   264 different workers worked during peak week and 125 workers worked during the pre-election week (see fns. 7, 8);  125 is less than 50 percent of 264 so the election could not be certified.

(f) Bonita and Wine World:  Utilized the margin of error method.   2.5 percent was found acceptable while 7 percent was found an unacceptable margin of error.   The margin produced in the current case would be 14 workers less than 50 percent of peak or a margin of error of 5 percent.   This would produce an unacceptable margin of error and the election could not be certified.

Our function is not to analyze the various methods used to determine peak and choose the one we like best.   The ALRB is the entity charged with implementing ALRA policies.   Its interpretation of policy, so long as it effectuates the purpose of the law, is entitled to deference.  (J.R. Norton Co. v. Agricultural Labor Relations Bd., supra, 26 Cal.3d at p. 29.)   However, the California Supreme Court has recognized the importance of judicial review by appellate courts to prevent arbitrary administrative action tending to diminish the vitality and integrity of the labor election process.  (Id., at p. 33.)   The Legislature's requirement of 50 percent of peak employment to trigger the election process would be eviscerated by formulae that fail to provide the maximum participation of affected workers.

 Inasmuch as none of the formulae previously utilized by the Board would result in a certification of the election in the instant case, we examine the hybrid formula utilized here to determine if it satisfies the statutory 50 percent of peak requirement and comports with the statutory purpose of maximizing employee participation in the election process.   We conclude it does not.

The basic defect in the combination of formulae employed in this case is that the average numbers utilized bore no relationship to the actual number of workers that were employed and thus would have been eligible to vote in the peak period or to the actual number of workers employed in the pre-petition payroll period and thus eligible to vote in the election.   The fundamental fact is that a worker is entitled to vote if he or she was employed anytime during the appropriate payroll period regardless of whether he or she worked one day or seven days.   Thus the average figure produced by totaling the number of workers each day and dividing by the number of days in the payroll period bears no significant relationship to the number of workers actually employed and entitled to vote.

The proposition cannot be better stated than was done by the IHE with Board approval in Donley Farms, Inc., supra, 4 ALRB No. 66:  “The legislature was attempting to insure that the actual voters in the election were of a sufficient number to be representative of all the workers in expressing their desire to have an election, regardless of outcome.   Since every agricultural worker who was employed during the pre-petition period is given the right to vote, I find no support within the Act for a proposition by which the value of a worker's interest in the election is dependent upon the number of days he or she works.   This is the inevitable result of averaging the employees in the pre-petition period.   The assumption is that if 6 different workers hold the same job position for a 7 day pay period because of high turnover in that position, they each are less representative of the wishes of the whole workforce during that time than one worker who holds a job position for an entire 7 days.   I find this result incompatible with the Act's stated purpose of ‘guaranteeing justice for all agricultural workers,’ and the legislative scheme under which each is given an equal vote, regardless of number of days worked.”  (Fn. omitted.)

Based on the figures disclosed in the record, some of which may not be entirely accurate (see fns. 7, 8, and 9, ante ), there were at least 264 different workers employed and thus eligible to vote in the peak period and some 125 different workers employed (see fns. 6 and 7, ante ) and thus eligible to vote in the pre-petition payroll period (see fns. 8 and 9, ante ).   125 is less than 50 percent of 264.   Nevertheless, by the use of the combination of formulae the IHE and the Board have concluded that petitioner's employment during the pre-petition period was well in excess of 50 percent of peak even if the Arizona workers were to be included.

There is yet another defect in the calculations utilized by the IHE and the Board in this case.   Using a divisor of three rather than seven for the contract workers resulted in an exaggeration of the average number of contract workers employed during the payroll period immediately preceding the filing of the petition.   It is undisputed the payroll period consisted of seven days, albeit a different seven days from petitioner's regular payroll period.   The latter problem, however, was remedied by the use of two separate averages, one for regular employees and one for contract employees.   There was no justification whatever for attempting to ascertain the average number of contract employees during the seven-day payroll period by dividing by three.   As was correctly pointed out in Donley Farms, supra, a worker is entitled to vote if he worked even one day in the appropriate payroll period.   More fundamentally, of course, as we have pointed out and as acknowledged in Donley Farms, the average number of employees during a payroll period bears no significant relationship to the number of eligible voters in a payroll period which is the meaningful statistic.

The Board has apparently belatedly recognized one or more of these defects in its calculations of peak and 50 percent of peak in the instant case, for at oral argument it advocated the use of yet another formula akin to that approved in Donley.   We agree that that formula would be appropriate, but as already explained, it results in a finding of 50 percent of peak or more only if the Arizona workers are excluded, and we have determined that they may not properly be excluded.

DISPOSITION

The ALRB's order certifying the election is annulled as it fails to comport with the policy of the ALRA in that it does not allow maximum participation of all affected workers and the calculation of peak employment was fixed arbitrarily and without justification.

The facts which gave rise to this controversy occurred over five years ago.   In light of fairness to all concerned, the election will be set aside “without prejudice to the right of any qualified party to file a new representation petition.”  (Franzia Brothers Winery (1981) 7 ALRB No. 14.)

Finally, in order to maintain stability in the agriculture labor field, this decision is to apply prospectively only.   Hence, this case has no effect on those elections which have previously been confirmed.  (See Wellenkamp v. Bank of America (1978) 21 Cal.3d 943, 953–954, 148 Cal.Rptr. 379, 582 P.2d 970.)

The order of the ALRB certifying the election of the UFW as the exclusive bargaining representative is annulled.

FOOTNOTES

1.   The preamble to the ALRA provides this legislation was designed “to ensure peace in the agricultural fields by guaranteeing justice for all agricultural workers and stability in labor relations [and] ․ to bring certainty and a sense of fair play to a presently unstable and potentially volatile condition in the state․”  (Stats.1975, ch. 1, § 1, p. 4013.)

2.   Unless otherwise indicated, all section references in this opinion are to the Labor Code.

3.   Under section 1144 the ALRB is authorized to make rules and regulations necessary to carry out the provisions of the ALRA.   Citations to the Administrative Code are those rules authorized under this provision.

4.   This section provides in part:  “․ In determining whether the showing of employee support submitted by the union is adequate, the regional director shall determine the average number of employee days worked on each day of the payroll period and shall compare names on authorization cards to names on the payroll to determine if sufficient cards signed by the employees whose names appear on the payroll have been submitted to equal at least a majority of the average number of employee days worked during the payroll period.”

5.   The results were as follows:  36 in favor of the UFW, 25 for no labor organization, and 3 unresolved challenged ballots.

6.   The IHE concluded there were 199 different persons in this group during peak week.

7.   The IHE concluded approximately 65 different persons made up this group during peak week.7/27/37/47/57/67/77/8(W)(Th)(F)(Sat)(Sun)(M)(Tu)California Melon& Farm WorkersEmployed6109108 71116 70106105 ArizonaDew Pickers7 39 39 35 37  0 45 57   Total148147106153 70151162

8.   The record is silent as to the total number of people in this group during the pre-election week.   Because these are “regular” employees and the maximum number working any one day was 15, it is reasonable to infer, and we do infer for purposes of decision, the total number of “regular” workers was 15.

9.   The IHE estimated 110 people made up this group during the pre-election week.11/511/611/711/811/911/1011/1111/12 (W)(Th) (F)(Sat)(Sun)(M)(Tu)(W)Regular8151512115 13 12Contract9 0 0 00 69 95 69  Total151512115 82107 69

10.   All figures were rounded to the nearest integer where necessary.

RICKLES, Associate Justice.

MORRIS, P.J., and KAUFMAN, J., concur.