Wales v. Jack M. Berry, Inc. , 192 F. Supp. 2d 1269 ( 1999 )

192 F. Supp. 2d 1269 (1999)

Gedeon WALES, et al., Plaintiffs,
v.
JACK M. BERRY, INC., et al., Defendants.

No. 95-66-CIV-FTM-23(B).

United States District Court, M.D. Florida, Ft. Myers Division.

March 25, 1999.

A. Neal Barkus, Jeffrey B. Hardie, Hunton & Williams, Washington, DC, D. Alan Rudlin, Hunton & Williams, Richmond, VA, Thomas M. Mackall, Hunton & Williams, Washington, DC, Cathy L. Lucrezi, Law Office of Cathy L. Lucrezi, Ft. Meyers, FL, for Plaintiffs.

G. ParisSykes, Jr., David P. Phippen, Kilpatrick Stockton, L.L.P., Atlanta, GA, James Lawrence Nulan, Henderson, Franklin, Starnes & Holt, P.A., Ft. Meyers, FL, for Defendants.

FINDINGS OF FACT and CONCLUSIONS OF LAW

WILSON, United States Magistrate Judge.

This is an action on behalf of a large number of citrus workers alleging violations of the Migrant and Seasonal Agricultural Worker Protection Act ("AWPA"), 29 U.S.C. 1801-1872, and the Fair Labor Standards Act ("FLSA"), 29 U.S.C. 201-219, during the three citrus harvest seasons from 1991-1994 at citrus groves owned and operated by the defendants. The non-jury trial in this case has been bifurcated between the issues of liability and damages. The first phase of the trial focused primarily upon the number of hours and days worked during the workweeks of the three harvest seasons. Other issues, however, were also addressed. Based upon the testimony that was presented, and the evidence that was adduced, the following findings of fact and conclusions of law are rendered pursuant to Rule 52(a), F.R.Civ.P.

A. THE PARTIES

1. The plaintiffs in Count I are approximately 2800 migrant and seasonal citrus pickers employed by the defendants to harvest citrus during the harvest seasons of 1991-1992, 1992-1993 and 1993-1994. Sixty-five of those pickers have opted into the representative action under the FLSA alleged in Count II. A small number of individual plaintiffs have alleged retaliation in Count III. In Count IV, some plaintiffs have alleged that the defendants failed to pay promised bonuses and incentives.

2. The overwhelming majority of the plaintiffs are originally from Mexico, Guatemala or Haiti. Most of the plaintiffs assert that they do not either speak or read English. The ability to understand English would obviously vary among such a large number of individuals.

3. Defendant Jack M. Berry, Inc., ("Berry") or its predecessor owned approximately 20,000 acres of orange, grapefruit and other citrus groves in Southwest Florida during the three-year period in question. The majority of these were located in and around LaBelle and Immokalee, Florida. During the 1991-1994 period, Berry was owned and controlled by Jack M. Berry, Sr., and his wife.

4. Berry contracted with Defendant Eagle Lake Harvesting Corporation ("Eagle Lake") to harvest citrus from its groves. During the period in question, Eagle Lake was owned entirely by Jack M. Berry, Sr., and his wife. All of Eagle Lake's income came from harvesting citrus from the Berry groves.

5. Berry and Eagle Lake were a single, integrated operation. The two companies operated from the same principal office building. The president of Eagle Lake, Michael Murphy, was an employee of Berry. He attended Berry management meetings and reported to Cal Sellers, executive vice president of Berry.

6. All of the officers, managers and other salaried personnel assigned to run Eagle Lake were employees of Berry. All of the vehicles and equipment used by Eagle Lake in its harvesting operations were owned by Berry.

7. Berry operated a juice processing plant at its LaBelle, Florida, headquarters. Much of the citrus harvested by Eagle Lake was processed at this plant. Eagle Lake's harvesting schedule was determined by the needs of the processing plant.

8. Eagle Lake operated with a three-tiered management structure. Reporting to Murphy was William Rollins, the harvest manager. Reporting to Rollins were two harvest supervisors. The harvest supervisor position was filled at pertinent times by Bernardo Barnhart, Pete Aguilar and Raul Rodriguez.

9. The harvest manager and supervisors were responsible for scheduling the groves for harvest, overseeing the harvest operations, inspecting the work of the harvesting crews, and giving instructions to crew leaders.

B. COVERAGE UNDER FLSA and AWPA

10. The defendants were aware that their operations were covered by the FLSA and that they were required to pay the plaintiffs at least the minimum wage for each hour worked in the workweek. The defendants mounted posters at the entrances to their groves and in their offices announcing to employees their right to receive the minimum wage, which at that time was $4.25 per hour. The defendants were also aware that the FLSA required Eagle Lake to keep accurate time records.

11. The defendants, in addition, were aware that their operations were covered by the AWPA.

C. EAGLE LAKE'S HARVESTING OPERATIONS

12. Berry's groves around LaBelle and Immokalee mostly were in remote, sparsely populated areas. They had no permanent structures. Trees were planted in bedded rows and were irrigated by ditches running parallel to the beds.

13. The defendants grew and harvested about 10 to 12 different varieties of oranges. The harvest of early season varieties began in mid-October to November. Mid-season oranges were harvested beginning in the last half of January. Valencias were the primary late season orange, and their harvest began in late March and could last into June.

14. Prior to the harvest, Eagle Lake hired as many as 50 to 60 workers at the minimum hourly wage to perform grove maintenance. Eagle Lake began hiring for the harvest from mid-October through early November. The size of the picking staff grew as workers returned to the area. The maximum number of pickers on the payroll at one time was between 350 and 400. Turnover in the picking crews was somewhere in the vicinity of 25%. Crews that remained at the end of the season were released at the same time. Those crews had a full day's work up until the season ended. When pickers were released, they were no longer employees of Eagle Lake, and they would not necessarily return from season to season.

15. The plaintiffs harvested citrus by hand. This was done either standing under the tree or by climbing a ladder to reach the fruit. Citrus trees have thorns. Consequently, pickers would normally wear some protection against the thorns, such as long-sleeved shirts and gloves. Gloves were not provided by the defendants.

16. Picking citrus is a difficult job, requiring strength, stamina and agility. It has been compared to an athletic ability. Accordingly, not every picker is equally skilled. Thus, the productivity of pickers on the same crew can vary. Further, the productivity of an individual picker can vary from day to day. The variations in productivity could be due to physical factors and they could also be due to motivational factors.

17. Fruit was normally placed into a large pick sack worn by the picker. When full, the sack weighs about 90 to 100 pounds and is the rough equivalent of the traditional "box." Actual boxes, however, are no longer used at this stage of the harvest.

18. After the sack is filled, the picker deposits the fruit in a large receptacle called a bin. The bins at Eagle Lake held 10 boxes of fruit. Typically two bins were assigned to each picker each workday. These would be positioned by the picker near the trees he or she was picking.

19. Full bins of fruit were lifted by, and dumped into, a motorized vehicle called a high-lift or "goat." The goat held eight bins of fruit.

20. When a picker's full bin of fruit was dumped into a goat, he or she was either given a token or a tally card that was punched by the goat driver. The crew leader would also keep a record of the number of bins of fruit lifted and dumped for each picker during the day. Bins that were filled one day but not lifted and dumped until the next day were not credited to the picker until they were dumped.

21. After the goat was filled, it was driven to a large trailer and dumped. When the trailer was full, it was picked up by a trailer driver and its contents were taken to the juice processing plant in LaBelle. Eighty to one hundred trailers of fruit would be processed at the plant each day.

22. When a goat broke down, the bins could not be emptied. The goat would have to be taken to the shop for repair or simply replaced. However, a spare goat was always available. If there were multiple breakdowns, equipment would be coordinated in an effort to keep the crews picking.

23. Nevertheless, there were times that lack of useable equipment would delay picking. Thus, a crew could arrive in the morning and find that the bins were still full from the day before, so that the pickers had to wait to begin working until the bins were emptied. Similarly, at times, empty bins were unavailable. Delays could also be caused by the absence of goats or empty trailers.

24. The plaintiffs presented testimony that these equipment-related delays were frequent and often lengthy. I find this testimony to be exaggerated and not credible.

25. When a crew finished picking its assigned grove, a supervisor had to inspect the grove for missed fruit, called "shiners," and for trash. Crews were forced to wait during this inspection process. If the grove had not been picked clean or if trash was found, the crew was required to remedy this. At times, the crews would have to wait for a supervisor to arrive. When a crew was cleared to begin picking another grove, it was transported to that grove during the workday by the crew leader.

26. I reject as not credible the testimony that crews often had to wait for several hours for an inspection to take place. It obviously was in the defendants' interest to expedite this process in order to keep the crews working.

D. THE CREW LEADERS

27. Eagle Lake engaged crew leaders to supervise the harvesting crews directly. "Outside" crew leaders employed their own crews. "Inside" crew leaders supervised pickers who were employed directly by Eagle Lake and used equipment and supplies provided by Eagle Lake. The plaintiffs were on crews supervised by inside crew leaders. Crew sizes varied; one witness said the crew size was approximately 15 pickers, while another stated that the size of his crew was between 20 and 30 pickers.

28. Inside crew leaders executed new contracts each year. These contracts, among other things, made the crew leaders responsible for the supervision of the pickers, for the keeping of time records, and for acting as paymaster. The same basic form contract was used from year to year.

29. Crew leaders were compensated on the basis of each box of fruit harvested by their crew and, accordingly, had an economic incentive to ensure that pickers on their crew worked diligently.

30. The inside crew leader contracts restricted the hiring and firing authority of the crew leaders, but, in practice, crew leaders hired and fired the pickers as they saw fit.

31. For the most part, the pickers were transported into the groves by the crew leaders. There was no time clock or set schedule. The crew leaders varied in the times at which they arrived in the morning and left at night. Thus, the crew leaders generally determined the length of the workday. However, the pickers would take breaks during the day and could even return to the transportation vehicle to wait for the others to finish work. In addition, pickers at times would not begin work until early-morning moisture on the fruit had dried.

32. The pickers were entitled to a lunch break of one-half hour, which was not compensated. The pickers, however, could work through the break, or could take a shorter, or longer, time for lunch.

33. The crew leaders were responsible for reporting, among other things, the pickers' hours. However, they were not able to constantly observe each picker. Consequently, the hours reported by the crew leaders are estimates.

34. The groves were open for the crews to work seven days per week. There was no evidence, however, of any minimum number of days per week that was required. The number of days that crews worked varied, and some pickers would work fewer days than other members of their crew.

E. THE PIECE RATE AND THE DEFENDANTS' GROVE QUALITY

35. The basic method for paying the plaintiffs for citrus picking was the piece rate. Michael Murphy set the piece rate that Eagle Lake would pay. Eagle Lake prepared disclosure statements that stated that the piece rate would be 65¢ per box "and up" for oranges and 40¢ per box "and up" for grapefruit. Oranges comprised approximately 80% of the harvest and grapefruit made up about 20%.

36. In the citrus industry, a piece rate can vary depending on the quality of a grove, and the difficulty of picking in the grove. The piece rate paid for oranges in the defendants' groves was mostly 65¢ per box even though the quality of the defendants' groves was mixed. However, piece rates were paid above that amount for picking oranges.

37. The Berry groves varied in quality. A majority of the Berry groves in LaBelle were planted in the 1960's and tended to be taller. Tall trees are more difficult and time consuming to pick. The irrigation ditches also created an awkward platform on which to place a ladder. It was estimated that 30% to 40% of the Berry groves were good picking, 30% were average and 30% to 40% were "tough picking."

38. Michael Murphy expressed a desire for workers that could pick one bin per hour. This was possible in the best of the defendants' groves, but not realistic in the groves that were of average or poor quality.

39. The piece rate for the good Berry groves was comparable to that paid at neighboring groves owned by others, but was below what was being paid by others for rough picking. This had an effect on the defendants' ability to attract labor. The top pickers—"the lions"—would go to another grove if more money could be made there.

40. During this period, harvest manager William Rollins suggested to the defendants that they raise the piece rates in their poor quality groves. This idea, however, was rejected.

F. THE DEFENDANTS' TIME-KEEPING PRACTICES

41. Crew leaders were assigned the responsibility of recording the hours worked by each picker. Time clocks were never used. Crew leaders were not instructed to record the time when crews entered and exited the groves, or the time when the pickers started and stopped work. Accordingly, the defendants have no daily, contemporaneous time records of the starting and ending time of the workday of any picker.

42. Crew leaders were instructed to record the total number of hours worked for each picker, instead of recording the beginning and ending time of the work day. However, the defendants have no daily, contemporaneous time records of the length of the meal periods taken by any picker, of the specific amount of time that was excluded from paid time for any picker for non-work activities, or of the specific amount of time any picker was required to wait due to factors beyond his or her control.

43. From November 1991 until March 1993, crew leaders recorded time on spreadsheet documents called time logs. All pickers on the crew were listed on the time log. When time logs were in use, more than a single day of work could be recorded on a single log.

44. After March 1993, crew leaders recorded time on cards the size of a parking ticket. These were called scan cards. There was a separate scan card for each picker on the crew. Only a single day of work could be recorded on the scan card.

45. Both the time log and the scan card required the crew leader to record the amount of fruit picked, either in boxes or bins, and the hours worked by each picker for the day. However, because, as indicated, the crew leaders could not directly supervise crews spread out over a broad area, the crew leader could not know the precise amount of time each picker worked.

46. Crew leaders filled out the time logs and scan cards after the work day ended, either that night or the following day. They usually did so in whole hour increments. At the time the time logs and scan cards were filled out, the crew leader knew the total number of full bins for each picker that had been lifted and dumped during the day.

47. No hours worked for a day were credited to a picker for that day if the fruit he picked was not lifted and dumped into the trailer on that day. Hours were credited, however, on the day that the fruit was dumped.

48. Pickers, as indicated, were allowed one-half hour for lunch each day. Eagle Lake treated this as unpaid time, and crew leaders were instructed not to count that one-half hour as time worked in the groves. The pickers were also given two 15-minute breaks each day during which they could relax. The pickers could choose when to take those breaks.

49. The plaintiffs contend that the defendants sought to pay only piece rate compensation to citrus pickers and that they accomplished this goal by understating the plaintiffs' hours. The plaintiffs assert further that a method used by the defendants to pay the piece rate, and not the minimum wage, was matching one hour for each bin picked.

50. To support this contention, the plaintiffs presented the testimony of Dr. David Peterson, an expert in applied mathematics. Dr. Peterson divided his study of the defendants' records into three periods, which did not entirely coincide with the three picking seasons at issue: the period from November 1991 through December 1992 (Period I), the period from January 1, 1993, through mid-December 1993 (Period II), the period from mid-December 1993 through June 1994 (Period III).

51. The time logs in evidence from Period I show that crew leaders at times matched hours to bins picked. Thus, a statistical analysis of the defendants' payroll data base reveals that the hours recorded exactly matched the number of bins picked 21.9% of the time.

52. Period II, under Dr. Peterson's analysis, began in late 1992. This corresponds with an announcement to the crew leaders that "there should be no falsifying or tampering with the records," and with the changeover to a new computer payroll system and the introduction of scan cards. Thereafter, some crew leaders stopped matching hours to bins picked and began recording hours in a way that more closely reflected the hours actually worked. Many pickers began to receive the minimum wage per hour instead of straight piece rate compensation.

53. Consequently, the median number of hours per week credited to pickers rose from 26 in Period I to 38.5 in Period II. This resulted in a decline in the median wage paid per hour during the period to $4.25, the minimum wage. In addition to being the median, $4.25 was overwhelmingly the most frequent rate of pay in Period II. The difference between the gross earnings per hour in Period I versus Period II was nearly 50 standard deviations, a difference far too large to be attributable to chance.

54. The defendants contend that the increase in the median hours recorded by crew leaders in Period II resulted from the false overreporting of hours by crew leaders. In support of this contention, the defendants point to instances where crew leaders apparently recorded for pickers more time worked than there was sunlight. There was also evidence of implausibly low production, such as picking one bin in eight hours. This evidence, however, does not demonstrate any widespread overstatement of hours. Significantly, the defendants have not advanced any substantial reason why the crew leaders would overstate hours on a large-scale basis.

55. The defendants also support their theory of overstated hours with the testimony of Dr. William Kerr, an expert in labor economics. Dr. Kerr noted that, if the hours in Period II were accurate, there would have been a remarkable drop in productivity during that period.^[1] However, as Dr. Kerr indicated, there was a labor shortage that period so that less skilled and adept workers may have been hired. He also pointed out that more frequent payment of the hourly wage would act as a disincentive to pick. Under these circumstances, the defendants cannot establish their contention of overstated hours by reliance on the drop in productivity.

56. The plaintiffs not only dispute the defendants' argument of overstated hours but assert further that hours continued to be understated. They have not, however, adduced any persuasive evidence of that assertion. Moreover, if there were underreporting there is no reason to think that it would amount to more than an offset of the crew leader's overstatements.

57. In sum, the records for Period II seem to reflect most accurately the hours worked picking in the Berry groves. Nevertheless, both sides have managed to raise questions about the reliability of the specific hours stated in those records.

58. Period III of Dr. Peterson's analysis starts in mid-December 1993. This period was a difficult financial time for the defendants, since revenues declined steeply, as compared with the prior period. There was too much fruit, and the prices were low. As a consequence, a number of employees were laid off in order to cut costs, including Tom Rogers, Berry's vice president for human resources.

59. In the midst of this financial difficulty, the defendants' executives, including Murphy, became upset at the amount of supplemental compensation having to be paid pickers to bring them up to the minimum wage. A report called a minimum wage report revealed those figures on a weekly basis. Murphy had also seen time records that showed crew members with a uniform number of hours worked but a widely different number of bins picked, which he believed was a sign that crew leaders were overstating hours. The records from this period also showed many workers having poor production.^[2]

60. Murphy angrily conducted a meeting with crew leaders. Among other things, Murphy told the crew leaders that they had been overstating the hours they recorded for the pickers. He said that this was like stealing and he wanted it stopped. Murphy indicated that he wanted to see pickers who could pick one bin per hour.

61. After the meeting, an unspecified number of crew leaders resumed matching hours to the number of bins picked in a day. The percentage of records in Dr. Peterson's Period III showing a one-to-one match between hours and bins picked was 22.4%, a figure higher even than in Period I. The number of hours credited per week in Period III declined, and the median gross wage per hour climbed above $4.25.

62. The number of times that minimum-wage supplemental compensation was paid to pickers was greatly lower in the months following January 1994 than in the same months of the previous year. In addition, Eagle Lake's payroll costs substantially declined in the 1993-1994 season in comparison to the season before.

63. The plaintiffs contend that Murphy instructed the crew leaders to match hours to bins. However, Murphy's angry statement that some pickers, in effect, were stealing from the company and his expressed desire for the hiring of people who could pick one bin per hour does not equate to an instruction that the crew leaders should falsely record one hour for every bin picked. Furthermore, the highest percentage of matching that has been shown is 22.4%. Thus, matching did not occur in 77.6% of the cases. If Murphy had issued instructions to the crew leaders to engage in matching, there is no reason to think he would have accepted such a high percentage of non-matching. Consequently, the evidence fails to persuade me that Murphy told the crew leaders to match hours to bins.

64. Some crew leaders may have perceived Murphy's request for workers who could pick one bin per hour as a direction to match hours to bins. Others simply may have taken that approach as the easiest method to report hours. In addition, some pickers may, in fact, have picked one bin per hour, so that there was not inaccurate matching, but a correct recording of hours. Therefore, the fact that there was some matching does not show that the matching resulted from Murphy's direction.

G. THE UNRELIABILITY OF THE DEFENDANTS' PAYROLL RECORDS

65. Throughout the three harvest seasons in question, the defendants maintained a computerized payroll data base that contained information relating to the plaintiffs' hours worked, days worked, fruit picked and weekly pay. The original sources of that information were the time logs and scan cards completed by the crew leaders. Because of the recording practices of the crew leaders, as well as the alteration of certain records in the harvesting office, the defendants' payroll data base is not reliable with respect to the hours worked.

66. Crew leaders turned in the time logs and scan cards to the Eagle Lake harvesting office, located at the Berry headquarters in LaBelle. During Period I the boxes picked and the hours recorded by the crew leaders were manually entered into the payroll system by clerks in the harvesting office. After scan cards came into use in March 1993, they were automatically read by a machine reader, which avoided the manual entry step.

67. The primary data entry employee in the harvesting office until the late summer of 1993 was Carole Terrell. She was trained by Tom Rogers, the director of human resources and later vice president for human resources, in the procedure for entering information from the time logs into the computer system.

68. The computer program would show an error message whenever the daily piece rate compensation for a picker was less than the hourly compensation earned for that day (the number of hours recorded by the crew leader times the minimum wage of $4.25). Terrell would reduce the number of hours entered one hour at a time until the message disappeared. Not only did Terrell reduce the hours recorded by crew leaders when entering them into the computer, she also altered the entry on the time log itself. Terrell trained several harvesting office clerks to reduce the hours recorded by crew leaders and alter time logs in the same method she used.

69. About one week into the 1992-1993 harvest season, Murphy noticed apparent changes in the time records and mentioned it to Rogers. Rogers went to Terrell and asked her if she were changing hours. When Terrell said that she was, Rogers told her to stop.

70. Rogers knew that the practice of changing time records in the office had extended back at least to the 1991-1992 season. However, at the time no effort was made by the defendants to determine which pickers had been underpaid during that season or to remedy the underpayment. The computer system used in 1991 and 1992 provided the error message on a daily, rather than a weekly, basis. However, the defendants' minimum wage supplement need only be provided on a workweek basis. Many of the changes resulting from the erroneous data entry practice had no impact on wages due for a workweek. After this action had been pending for several years, the defendants paid into the court registry $1,908.77 in an attempt to remedy the underpayment.

71. The plaintiffs contend that the defendants' employees, particularly Rogers, intentionally trained Terrell to alter the data. This contention is contradicted by the fact that Murphy and Rogers spontaneously ended the practice when they discovered it. Consequently, I find that Terrell's methods resulted from a misunderstanding of her training.

72. Nevertheless, regardless of the cause of the alterations of hours during the 1991-1992 season, the fact remains that they did occur and they render the time records for that season unreliable. Further, the extensive matching that took place during the 1993-1994 season makes the time records for that season unreliable. And with respect to the 1992-1993 season, the parties, while not establishing their respective contentions that hours were underreported or overstated, did demonstrate that those records were suspect concerning hours. However, the records of hours for that season do offer some supporting information for estimating the hours worked.

73. On the other hand, the defendants' records provide an adequate basis for determining the number of days worked by a picker in a workweek. By April 1993, crew leaders were to fill out daily scan cards to record pickers' work activity. Those cards required the crew leader to mark the day of the week upon which the work took place. There was no testimony showing that there was any significant failure to record on the scan card the specific day worked. The scan cards therefore would establish the number of days worked for each workweek after their use commenced in April 1993.

74. Prior to the use of the scan cards, work was recorded on a time log. Usually, a time log contained a single day of work for each person on a crew that had worked that day. However, some time logs were for multiple days and reflected data for any picker in the crew who worked one or more of the multiple days listed on the time log. Generally, when the time log included data aggregating picking work done on multiple days, it is not possible to tell from the time log whether any individual picker listed on the time log worked only on one of the multiple days, on more than one but not all of the multiple days, or on all of the multiple days aggregated on the time log.

75. In the vast majority of the time logs, the number of days worked per week can be determined from those records. In the instances where they cannot because of the apparent recording of multiple days on one time log, the number of days worked per week can, and should, be estimated through the method employed by the defendants' witness Amit Mehta.

H. THE CALCULATION OF THE HOURS WORKED

76. Because of the unreliability of the defendants' records regarding hours, it is necessary to estimate the number of hours that the plaintiffs worked. The plaintiffs contend that over each of the three seasons they averaged 11 hours of compensable working time per day. They supported this contention with the testimony of several pickers, as well as a few crew leaders. One picker said he worked from between 6:00 and 7:00 A.M. until 8:00 and 9:00 P.M.; two stated that they worked 12 to 13 hours per day; two testified that they arrived at 6:00 A.M. and left at 6:00 PM., and another said he worked from 5:00, 6:00, or 7:00 A.M. until dark, which was 10 to 12 hours.

77. I credit this testimony only to the extent that it indicates that, in general, the pickers worked a full day. However, I reject the testimony as exaggerated and thus not credible to the extent that it purports to state the hours that the pickers worked. The reasons for the rejection include the following: (1) the plaintiffs' testimony was tainted by a self-interest; (2) portions of the testimony were inherently incredible (e.g. their alleged working hours exceeded the hours of sunlight), while other portions were implausible; (3) the purported working hours substantially exceeded hours recorded in objective time studies; (4) the claimed hours were much greater than the fairly accurate average of 8.3 hours reflected in the defendants' records during Period II; and (5) the stated hours are belied by the pickers' productivity, especially in Period II. Furthermore, plaintiffs' counsel apparently do not fully accept much of the testimony at issue, since their contention of an average of 11 hours work per day is below the number of hours asserted by at least a majority of the pickers who testified on the subject.

78. The evidence establishes that the average number of hours worked per day over the three seasons is not greater than nine. Surveys done during the relevant time frame under the authority of the federal and state governments in order to determining prevailing wages in the citrus harvesting industry provided objective information concerning the average number of hours worked per picker in a day in the Southern Region of Florida, which includes the Berry groves. The surveys provided the following results:

The 1991-1992 Season:
Early and Mids      6.18 hours per day
Valencias           7.73 hours per day
The 1992-1993 Season:
Early and Mids      8.18 hours per day
Valencias           8.76 hours per day
The 1993-1994 Season:
Early and Mids      7.82 hours per day
Valencias           6.89 hours per day

79. The plaintiffs object to these results on the ground that the surveys were not shown to have included adequate inquiries about compensable waiting time. This speculative objection, however, does not undermine the basic validity of the surveys, since no evidence has been presented to indicate that any significant amount of compensable time was omitted. Nevertheless, to account for that possibility, the average of nine hours of work per day will be used. That amount is higher than all of the survey results, and more than one hour higher than four of the six survey findings.

80. An estimated average of not greater than nine hours per day is also supported by the defendants' records from Period II. Those records reflect an average of 8.3 hours per day. While the parties have demonstrated that the Period II records are not sufficiently reliable to be a basis for determining the hours worked for each of the individual pickers, they have failed to show that, as a whole, the records either underreport or overstate the average number of hours worked.^[3]

81. The defendants, in addition, presented testimony from three experts in the citrus industry. Their testimony was that the average or typical work day for a picker was six to eight hours. In general, I find this testimony, which was based upon years of experience, to be credible.

82. On the other hand, there was evidence, aside from the pickers' discounted testimony, that the pickers at Berry regularly worked more than eight hours. Thus, William Rollins, Berry's harvest manager from 1992 to 1994, testified that pickers normally arrived at the groves about 7:00 A.M. and during the winter months stayed until about 5:00 P.M. As the days lengthened, so did the time the pickers remained in the field. Rollins indicated that by June the pickers were in the groves for 12 to 13 hours. Rollins' testimony, however, while shedding light on the pickers' normal workday, did not purport to estimate the average number of hours worked per day during the three seasons in question. Consequently, I give greater weight to the combined evidence from the prevailing wage surveys, the defendants' records concerning Period II, and the citrus industry experts.

83. The defendants operated in a fluid and competitive labor market. Accordingly, they could not maintain an adequate labor force if they, as the plaintiffs assert, paid less than competing businesses and required the pickers to work significantly longer hours than the competing businesses.

84. Under all the circumstances, I find that nine hours is the appropriate estimate of average time worked by all pickers (from the lions to the lazy) over the three seasons. I give primary weight to the prevailing wage surveys and the defendants' records from Period II. This evidence is supported by the testimony of the three citrus industry experts, which I credit. The information from these sources indicates that an average eight-hour workday is at the top end of the range of time spent in the groves by pickers. Nevertheless, in light of the testimony of William Rollins, and the defendants' failure to maintain accurate records, which justifies giving the plaintiffs the benefit of the doubt (see infra, p. 1291), I am adjusting the average upward to nine hours per workday. This adjustment is sufficient to cover any waiting time spent by the pickers in the groves for which they were entitled to be compensated. If the estimated average workday of nine hours is off the mark, it is off the mark on the high side, but that result is justified by the defendants' failure to maintain accurate records.^[4]

I. ALLEGATIONS OF RETALIATION CONCERNING THE ALANIZ CREW

85. Early in the 1993-1994 season, several migrant pickers were recruited for the crew of crew leader Jose Alaniz after allegedly receiving oral promises from a person named Pedro of the minimum wage for 10 hours of work per day, six days per week. The particular plaintiffs assert further that, for a time, they got the amount of work and the pay they were promised, but were unable to pick many oranges because of the quality of the groves to which they were assigned.

86. The plaintiffs claim further that, after about two weeks of picking oranges, several of the crew members noticed that their pay had dropped. They complained to Alaniz, who first said he would straighten it out, but later said he could not do anything about it.

87. Fifteen to twenty pickers decided to go as a group to the harvesting office to complain. Among these were Francisco Tecum-Chiguil, Daniel Tamiac, Oscar Chiguil, Marcos Arriaga, Santiago Chiguil, Francisco Miguel Francisco, Santos Silverio Teguil and Pascual Son. At the office they complained to a secretary. The secretary informed the group that pay was based on piece work, not hours. The group then left.

88. Dissatisfied with that payment arrangement, some of the complainants stopped working for the defendants immediately. Others, however, continued to work for several days.

89. The members of the Alaniz crew were not fired by the defendants. Moreover, there is no probative evidence of any retaliatory activity taken by the defendants as a result of the complaint to the secretary.

J. THE TERMINATION OF DIEUJUSTE AND HIS CREW

90. One of the crew leaders working for the defendants during the 1993-1994 season was Madsene Dieujuste. His paperwork was a subject of complaints by women in the office. Rollins testified (Tr. I, p. 111):

He wouldn't match. In other words, he was writing down however. If they picked one tub or ten tubs, he would write down 10 tubs and how many hours they were in the field.^[5]

Murphy, consequently, instructed Rollins to tell Dieujuste that he could correct his hours or he could leave. In late February 1994, Rollins went to speak with Dieujuste in the groves and told him to balance his time log, or Rollins would have somebody drive him off the property, in which event he would be fired. Dieujuste responded, "Go get the guy" (id.).

91. Dieujuste and all the members of his crew were driven off the property on February 23, 1994. The members of Dieujuste's crew depended on the transportation he provided to get to and from work. Nobody from the company came to pick them up for work thereafter.

92. The crew members, as distinct from Dieujuste, were not fired by the defendants. It is true that the paramount disagreement with Dieujuste appears to have been his crew's failure to perform adequately. Nevertheless, the crew members could have returned to work for the defendants under another crew leader.

93. There is no probative evidence that Dieujuste's firing had anything to do with complaints about pay made to him by his crew members. More importantly, there is no probative evidence that the end of his crew members' employment relationship was caused by any complaints they made to him. Accordingly, I find that the plaintiffs have totally failed to prove that the end of their employment was in retaliation for complaints about pay.

K. THE PAYCHECK STUBS AND PIECEWORK CHECK ATTACHMENTS

94. Pickers were paid by check each Friday. Checks delivered to pickers on Friday were not for the workweek then ended but for the preceding week. Checks were generated in the payroll office at Berry and delivered to Eagle Lake's harvesting office for separation into groups by crew leader. Checks were then delivered to the crew leaders for distribution to the pickers. The check could be detached from the check stub at a perforation. The checks and the stubs were printed only in English throughout the period in question. The check stubs for the 1991-1992 season and the beginning of the next season contained such information as the number of hours worked, the number of boxes picked, the net pay and Eagle Lake's employer identification number.

95. At the beginning of 1993, the defendants began using new computer payroll software. That computer program generated a document called a piecework check attachment, which was intended to contain information required by law to be provided to the pickers at the time of payment. The attachment, in fact, was not attached to the check or put into a pay envelope.

96. However, the attachments were sorted in the office and laid upon the checks for distribution to the crew leaders, who in turn gave them to the workers. I find that, in general, the attachments were provided to the workers. I reject as not credible testimony offered by the plaintiffs that the attachments were rarely, if ever, distributed. There was no probative evidence adduced that established specific occasions when the attachments were not provided.

97. The attachment was only issued in English. It showed the number of boxes picked and the piece rate that was applicable. The check stub continued to indicate the hours worked and the net pay. However, there did not appear on either document an employer's identification number.

L. THE INCENTIVE PROGRAM

98. In an attempt to reduce turnover, the defendants instituted a bonus program under which each picker who was on the payroll before a certain date and who stayed until the end of the season would receive $100. In addition, for every box picked above a specified quantity, a bonus of 10¢ would be paid, with 5¢ to be paid during the current pay period, and 5¢ to be deferred and added to the signing bonus. In order to remain eligible for the signing bonus, a picker could not take off more than three consecutive days unless an acceptable reason was given.^[6] In addition, a picker who was not on the payroll at the end of the season because he had been fired would not get the bonus.

99. Several plaintiffs, including Eustaquio Calderon and Silvestre Galvan-Castro, who worked until the end of the picking season in June 1993, allege that they did not receive their bonus. Not only did Calderon and Galvan-Castro fail to testify, but the defendants have pointed to evidence showing that each missed three consecutive days without apparent excuse. Consequently, these two have failed to prove their claim. Further, the plaintiffs' post-trial proposed findings of fact and conclusions of law did not even refer to any pickers who had alleged a failure to receive bonus or incentive payments for the 1992-1993 season.^[7]

100. The defendants continued their bonus and incentive program in the 1993-1994 season. The plaintiffs claim that the defendants improperly denied Santiago Chiguil, Francisco Tecum-Chilguil, Francisco Miguel-Francisco, Serapio Tecum, Matias Morales, Pascual Son, Marcos Arreaga and Santos Silverio-Tiguila the opportunity to earn the bonus and incentive by terminating them in retaliation for complaining about their pay in December 1993. It has previously been found that the defendants did not engage in a retaliatory discharge of these workers. Since these plaintiffs did not stay until the end of the season, they were not entitled to any bonus or incentive payments.

101. A similar claim was made by members of the crew of Madsene Dieujuste: Gedeon Wales, Fritz Dorvil, Louis Absolu, Regisma Etienne, Buliva Celeste, Bernadette Miclisse, Marcelo Vega Chavez, Dumacier Bonhomme, Admetton Torchon, Marie Cadet, Eddy St. Jean, Daniel Sanon, Maccia Cadet, Edner Joseph, Florencia Blanc, Dinel Chevry, Claudette Charles, Ancelio Phanor, Loruis German, Prevoir Polissaint and Jacques Chevry. Since those workers' claim of retaliatory discharge has been rejected, their failure to work until the end of the season disqualifies them from the bonus and incentive payments.

II. CONCLUSIONS OF LAW

A. JURISDICTION and STATUS OF PARTIES

1. Jurisdiction is conferred on this court by 29 U.S.C. 216(b) for the claims brought under the Fair Labor Standards Act (FLSA), by 29 U.S.C. 1854(a) for the claims brought under the Migrant and Seasonal Agricultural Worker Protection Act (AWPA), and by 28 U.S.C. 1331, which governs cases involving federal questions. Leach v. Johnston, 812 F. Supp. 1198, 1205 (M.D.Fla.1992).

2. At all times relevant to this action, each plaintiff was either a migrant or seasonal agricultural worker within the meaning of the AWPA. 29 U.S.C. 1802(8)(A), 1802(10)(A). Thus, while the defendants asserted in the pre-trial statement that this matter was an issue of fact to be litigated, they presented no evidence on the subject. Further, they have made no argument on this issue.

3. Throughout the period they were employed by the defendants, the plaintiffs were engaged in the production of goods for sale in interstate commerce and were thereby employees covered by the minimum wage provisions of the FLSA. 29 U.S.C. 206(a).

4. At all times relevant to this action, both defendants were "agricultural employers" within the meaning of the AWPA. 29 U.S.C. 1802(2); Leach v. Johnston, supra, 812 F.Supp. at 1206. Similarly, both defendants were employers within the meaning of the FLSA. 29 U.S.C. 203(d), (g); see Saintida v. Tyre, 783 F. Supp. 1368, 1373 (S.D.Fla.1992).

5. The plaintiffs contend that the "single employer" test, first developed under the National Labor Relations Act and now used in cases involving other employment statutes, is appropriate for determining whether two otherwise separate business entities should be treated as one for analyzing acts and their potential liability for those acts. McKenzie v. Davenport-Harris Funeral Home, 834 F.2d 930 (11th Cir.1987); Bruce v. S & H Riggers and Erectors, Inc., 732 F. Supp. 1172 (N.D.Ga. 1990). If this test does apply to this type of action, it would certainly fit the facts of this case in view of the common ownership, common control and the interrelationship between the two corporations.

6. It is unnecessary to determine the applicability of the "single employer" test because, in any event, the defendants would still be "joint employers" of the plaintiffs. Antenor v. D & S Farms, 88 F.3d 925, 929 (11th Cir.1996); 29 C.F.R. 791.2, 500.20(h)(4)(i), 500.20(h)(4)(ii). Significantly, the defendants do not argue otherwise, or make any contention that the liability of the two corporations could differ.

B. AWPA STATUTE OF LIMITATIONS

7. Counts I, III and IV are brought under the AWPA. The plaintiffs contend that, since the AWPA does not specify any limitations period for private rights of action, this court should "adopt a state time limitation as federal law that is not inconsistent with federal law or policy." Clark v. Coats & Clark, Inc., 865 F.2d 1237, 1241 (11th Cir.1989). Arguably, it would be more appropriate to borrow a limitations provision from another federal statute, namely the FLSA. See DelCostello v. International Brotherhood of Teamsters, 462 U.S. 151, 161, 103 S. Ct. 2281, 76 L. Ed. 2d 476 (1983). The defendants, however, do not press for the borrowing of a federal limitations period; to the contrary, in the pre-trial statement the defendants asserted that Florida's two-year limitations period in § 95.11(4)(c), Fla. Stat., for an action to recover wages was applicable. In light of the defendants' lack of opposition, a state limitations provision will be used. See Barajas v. Bermudez, 43 F.3d 1251 (9th Cir.1994).

8. The plaintiffs contend further that their AWPA claims are most analogous to one of two state law actions in Florida: (1) an action founded on statutory liability; and (2) an action for breach of an oral contract, both of which are governed by a four-year statute of limitations. See Fla. Stat. § 95.11(3)(f) (statutory liability), Fla. Stat. § 95.11(3)(k) (oral contract). Importantly, the defendants in their post-trial submissions do not challenge this contention and do not make any argument in support of the use of the two-year period for wage actions. The defendants therefore will be deemed to have abandoned any contention that the two-year provision should apply.

9. Accordingly, the plaintiffs' AWPA claims will be treated as being subject to a four-year statute of limitations. Moncevoir Hyppolite v. Gorday, 1990 WL 80684, at * 10 (S.D.Fla.1990); Marquis v. United States Sugar Corp., 652 F. Supp. 598, 602 (S.D.Fla.1987).

C. AWPA VIOLATIONS

10. During the trial of this matter, the plaintiffs proved that at some point during the three seasons the defendants committed certain violations of the AWPA and its attendant regulations. The extent of those violations will not be established until the second phase of the trial has been completed. Some alleged violations, however, were not proven.

11. First, by failing accurately to record the hours worked by the plaintiffs, defendants violated 29 U.S.C. 1821(d)(1) and 1831(c)(1). Defendants' records were inaccurate as a result of the alteration by office personnel of hours worked and the failure by crew leaders to report correctly the hours worked.

12. Further, by failing to state accurately on the weekly pay documents given to the workers the hours worked by the plaintiffs, defendants violated 29 U.S.C. 1821(d)(2) and 1831(c)(2). These violations seem essentially to replicate the preceding ones.

13. The plaintiffs contend that by failing to provide them with information showing the number of boxes picked or the piece rate that applied to those boxes, and by otherwise providing confusing and misleading information to the plaintiffs, defendants violated 29 U.S.C. 1821(d)(2) and 1831(c)(2). This contention is apparently based on the allegation that the piece rate attachment was not regularly distributed.^[8] The piece rate attachments did contain the number of boxes picked of a particular type of fruit and the piece rate for that fruit. Since it has been found that the attachments were regularly distributed and that the plaintiffs have failed to prove any specific occasion when they were not, the plaintiffs did not sustain their burden of showing statutory violations of §§ 1821(d)(2) and 1831(d)(2) based upon a failure to provide an itemized statement each pay period containing piece rate information.

14. The plaintiffs also contend that, by providing wage statements only in English to migrant and seasonal agricultural workers who are not fluent in English, defendants committed per se violations of 29 U.S.C. 1821(g), 1831(f) and 29 C.F.R. 500.78. The statutory provisions, however, do not require weekly wage statements to be in some language other than English. While the statute directs that certain recruiter disclosure requirements and employer posting requirements, "as necessary and reasonable," must be provided in Spanish or other language common to the workers, that statutory mandate does not extend to the weekly wage statements. 29 U.S.C. 1821(g), 1831(f). Although the implementing regulation, 29 C.F.R. 500.78, arguably could be construed to cover the weekly wage statements, it is clearly more reasonable to interpret the regulation as being congruent with the statute. Accordingly, the defendants did not violate § 1821(a) or § 1831(f) because the weekly wage statements were not in Spanish or Creole.

15. The defendants violated 29 C.F.R. 500.80(d) by failing to include the employer identification number on wage statements provided to workers from January 1993 through the end of the 1993-1994 season. The regulatory requirement is not satisfied simply because the workers may have subsequently obtained the information from another document, such as a W-2 form.

16. At times the defendants violated 29 U.S.C. 1822(a) and 1832(a) by failing to pay the plaintiffs at least the minimum wage of $4.25 per hour per weekly pay period when due. The defendants have admitted that there were some weeks during the 1991-1992 season when the minimum wage was not paid. Whether there were additional violations cannot be determined until the second phase of the trial is completed.

17. By failing to pay the plaintiffs at least the minimum wage, the defendants also contravened 29 U.S.C. 1822(c) and 1832(c), which prohibit violating terms of a "working arrangement." The defendants created a "working arrangement" within the meaning of 29 U.S.C. 1822(c) and 1832(c) by posting an official Department of Labor poster that notified workers of their right to receive at least $4.25 for each hour worked in the workweek, and then, as they concede, violated that working arrangement without justification on at least some occasions by failing to pay the minimum wage.

18. Certain plaintiffs contend that the defendants also created a working arrangement regarding a "sign-on" bonus of $100 to be paid if a picker worked until the end of the season, and then violated that working arrangement by subsequently creating new attendance requirements. It has been found, however, that the plaintiffs failed to prove that the attendance requirements were not part of the arrangement throughout the period.

19. The plaintiffs alleged that, in the 1993-1994 picking season, the defendants improperly discharged two groups of plaintiffs in retaliation for complaining about their pay. It has been found, however, that each group has failed to prove their claims. Accordingly, those claims should be dismissed.

D. INTENTIONAL NATURE OF AWPA VIOLATION

20. Damages under the AWPA are limited to those resulting from intentional violations of the statute or its regulations. 29 U.S.C. 1854(c). In this context, however, "intentional" refers only to the civil standard of holding a party responsible for the natural consequences of its acts. Thus, the plaintiffs need not prove a specific intent to violate the law. Bueno v. Mattner, 829 F.2d 1380, 1385 (6th Cir.1987), cert. denied, 486 U.S. 1022, 108 S. Ct. 1994, 100 L. Ed. 2d 226. It is enough to show that the defendants deliberately and consciously engaged in the action that led to a violation. Campbell v. Miller, 836 F. Supp. 827, 830 (M.D.Fla. 1993).

21. Under this lenient standard, most, but not all, of the defendants' AWPA violations were intentional. The failure to keep accurate records of hours worked, and the related failure to state accurately the hours worked on the weekly pay documents, is not based, contrary to the plaintiffs' contentions, upon direct instructions from the defendants' management. See, pp. 1278-79, 1279-80, supra. However, some of the inaccuracies resulted from the changes deliberately made by defendants' office personnel. This conduct was clearly intentional within the meaning of § 1854 and thus supports an award of damages.

22. In Period III, there was substantial matching of hours to bins resulting in inaccurate records and wage statements. While it has been found that the crew leaders were not directed by management to match hours to bins, some crew leaders engaged in that practice following Murphy's angry speech, either because they perceived Murphy to be directing them to do so, or it was an easy method to escape Murphy's wrath.^[9] Murphy was monitoring the weekly minimum wage reports and could therefore see the change in reporting that took place after his speech. Furthermore, Murphy had to be aware of the implications of his speech. Most significantly, Eagle Lake's office manager testified that not only did she notice that hours would match bins on the scan cards turned in by the crew leaders, but also that she thought that crew leaders would come into the office to match hours to bins. Based upon the evidence, I am persuaded that the defendants knew that matching was occurring in Period III. In light of this knowledge, the violations of inaccurate records and wage statements that resulted from the matching in Period III were intentional and therefore support an award of damages.

23. The defendants also violated the AWPA by failing to pay minimum wages when due and, correspondingly, by failing to comply with the working arrangement to pay minimum wages. Under the civil standard that a party is responsible for the natural consequences of its acts, it would seem that the mere failure to pay the minimum wages that were due would constitute an intentional violation. In any event, it has been found that the defendants knew of certain situations where hours were not being recorded accurately with the result that minimum wages were not paid. Under this circumstance, the failure to pay minimum wages was intentional and thus a basis for an award of damages.

24. The defendants violated the AWPA's regulation when, after their new computer program was installed at the beginning of 1993, an IRS employer's identification number was not included on the weekly wage statements. This omission was clearly an inadvertent oversight. Significantly, the defendants had nothing to gain by the absence of an employer identification number and thus had no reason to omit the number. Under these facts, the violation cannot reasonably be said to have been done "intentionally." An award of damages for this infraction would grossly ignore the language used by Congress and constitute a form of strict liability. The plaintiffs therefore may not recover under § 1854 for the lack of an employer identification number on the weekly wage statements.

E. AWPA STATUTORY DAMAGES

25. The AWPA provides that a court may make an award of actual damages or an award of statutory damages of up to $500 for each violation of the Act. 29 U.S.C. 1854(c)(1). The amount of any such damages shall be deferred to the second phase of the trial. At that time, evidence may be presented on the factors that influence an award of statutory damages. The class action plaintiffs seek actual damages for the AWPA violations involving the failure to pay the minimum wage when due and the failure to meet the terms of the working arrangement. If this means that the plaintiffs are claiming the same type of unpaid wages that are normally awarded under the FLSA, they have failed to this point to provide any basis for such an award. That approach would seemingly render provisions of the FLSA superfluous. Consequently, the court in Martinez v. Shinn, 992 F.2d 997 (9th Cir.1993), for example, awarded unpaid minimum wages under the FLSA and statutory damages for that violation under the AWPA. Thus, absent persuasive authority, unpaid minimum wages will not be awarded under § 1854.

F. AWPA INJUNCTIVE RELIEF

26. In addition to damage awards, the AWPA specifically provides this court with the authority to impose "other equitable relief." 29 U.S.C. 1854(c). The plaintiffs make a number of specific proposals for permanent injunctive relief. Those proposals, however, have never been addressed. Consequently, the issue of whether injunctive relief is appropriate in this case and, if so, what that relief should be, will be deferred until the second phase of this case.

G. FLSA VIOLATIONS

27. The minimum wage provision of the FLSA requires an employer to pay its employees at least the minimum wage for each hour worked in a workweek. 29 U.S.C. 206. During the relevant time periods, the minimum wage rate in effect was $4.25 per hour.

28. A "workday" commences at the time the employee reports for work "in accordance with the employer's requirement, even though he is not able to commence performance of his productive duties until a later time." 29 C.F.R. 790.6(b).

29. For purposes of the FLSA, an "hour worked" includes time spent in active mental or physical labor for the employer, as well as time a farm worker spends waiting for a work assignment in the field. 29 C.F.R. 785.14; see also 29 C.F.R. 778.223, 785.15, 785.16.

30. Compensable hours worked within the meaning of the FLSA include, among other things, short rest periods, travel time between work sites during the work day, and time during which employees are required to be present at the work site but must wait to work or must wait to satisfy the procedures imposed by the employer (e.g., inspection of groves). 29 C.F.R. 778.318, 785.14, 785.15, 785.16, 785.18.

31. While some meal periods may be compensable, bona fide meal periods are not. 29 C.F.R. 785.19. The defendants allotted the workers 30 minutes for lunch, and the workers were not compensated for that time. There was testimony that at times some took shorter lunch breaks or, somewhat remarkably, ate while working. This testimony was not sufficiently specific to identify situations where there was something less than a bona fide lunch break. Further, the evidence was not detailed, probative, or credible enough to warrant the conclusion that, as a regular practice, workers did not take a bona fide lunch break. Consequently, the plaintiffs are not entitled to compensation for the customary 30-minute lunch time.

32. Generally, waiting time is compensable if the time predominately benefits the employer and cannot be effectively used by the employees for their own purposes. Mireles v. Frio Foods, 899 F.2d 1407, 1411 (5th Cir.1990). The time spent by the plaintiffs waiting for bins to be emptied, trailers to be provided, inspections to occur, or for other equipment to be made available is all compensable waiting time. Accordingly, such waiting time was taken into consideration in arriving at the estimated workday of nine hours.

H. PIECE RATE COMPENSATION

33. An employer may pay its workers a piece-rate wage, as long as the piece rate compensation paid each week, when divided by all hours worked, is at least the equivalent of the minimum wage. United States v. Rosenwasser, 323 U.S. 360, 364, 65 S. Ct. 295, 89 L. Ed. 301 (1945).

34. Count II of the Amended Complaint is a representative action under the FLSA on behalf of 62 plaintiffs which alleges that the defendants willfully failed to pay the plaintiffs the minimum wage for work performed in the three harvest seasons at issue. The plaintiffs have elected to seek liquidated damages for these FLSA violations.

35. To prove a violation of the FLSA, a plaintiff employee bears the initial burden of proving that he performed work for which he was not properly compensated. Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680, 66 S. Ct. 1187, 90 L. Ed. 1515 (1946). Where, as here, a defendant fails to keep accurate records relating to the number of hours that each plaintiff worked, a plaintiff's burden of proving damages is reduced: "It is enough under these circumstances if there is a basis for a reasonable inference as to the extent of damages." Id. at 688, 66 S. Ct. 1187. A plaintiff can satisfy this burden through testimonial evidence alone. See, e.g., McLaughlin v. Stineco, Inc., 697 F. Supp. 436, 451 (M.D.Fla.1988). Nevertheless, the absence of reliable records does not require a court to accept unquestioningly such testimony. Marshall v. Hope Garcia Lancarte, 632 F.2d 1196 (5th Cir.1980). As previously explained, the workers' estimate of their work-hours was not credible. A consideration of all the evidence established, rather, that an average workday was, contrary to the plaintiffs' testimony, no more than nine hours. Such an estimate, based upon the totality of the evidence, comports with Anderson v. Mt. Clemens Pottery Co., supra, and Marshall v. Hope Garcia Lancarte, supra.

I. LIMITATIONS PERIOD FOR FLSA CLAIMS

36. The defendants contend that a number of the FLSA claims are barred by the statute of limitations. The pertinent provision, which is contained in the Portal-to-Portal Act, bars claims for unpaid minimum wages if not commenced within two years, "except that a cause of action arising out of a willful violation may be commenced within three years after the cause occurred." 29 U.S.C. 255(a).

37. An employer willfully violates the FLSA where it either "knew or showed reckless disregard for the matter of whether its conduct was prohibited by the statute." McLaughlin v. Richland Shoe Co., 486 U.S. 128, 133, 108 S. Ct. 1677, 100 L. Ed. 2d 115 (1988). The Supreme Court explained further in Richland Shoe Co. that "the word `willful' is considered synonymous with such words as `voluntary,' `deliberate,' and `intentional.'" Id. at 133, 108 S. Ct. 1677.

38. The deliberate changes made by defendants' office personnel were previously concluded to have been done intentionally. Unpaid minimum wages resulting from those deliberate changes constitute willful violations.

39. Further, it has previously been determined that the matching that took place in Period III was known to the defendants and was therefore done intentionally. Accordingly, any violations resulting from matching in Period III were willful.

40. For those willful violations, the appropriate limitations period is three years. 29 U.S.C. 255(a). The plaintiffs, however, have not proven any other willful FLSA violations. Consequently, if there are additional FLSA violations, the limitations period for those violations is two years.^[10]

J. DEFENDANTS' RETALIATION

41. Retaliation against those who exercise their rights under the FLSA or the AWPA is unlawful. 29 U.S.C. 215(a)(3), 1855(a). Complaining to one's employer or requesting a pay increase consistent with statutory requirements qualifies as exercising a protected right. EEOC v. White & Son Enterprises, 881 F.2d 1006, 1010-11 (11th Cir.1989).

42. In this case, two groups of employees have asserted claims of an unlawful retaliatory discharge. As previously explained, however, the plaintiffs have failed to prove their claims.

NOTES

[1] Dr. Kerr, like Dr. Peterson, divided the three years into three periods that did not coincide with the harvest seasons. His periods also did not coincide exactly with Dr. Peterson's periods. References to Periods I, II and III will be to Dr. Peterson's periods.

[2] The poor production could not be blamed on the quality of the groves because some workers in the same grove remarkably out-performed their co-workers. See, e.g., Plt. Exs. 600, 603.

[3] It is noted that the plaintiffs and, to a lesser extent, the defendants treat all three seasons as having similar working conditions. Thus, for example, no meaningful effort was made to account for the substantial drop in productivity in Period II. Consequently, under the parties' approach, the average hours worked in Period II would apply to the other two periods.

[4] It is noted that any matching of hours to bins at nine or more would, in all likelihood, have no effect on the pickers' compensation. Thus, if there were matching of nine hours for nine bins, a picker would have to work a full 14 hours before his minimum hourly wage would exceed his compensation at the lowest piece rate for oranges of $6.50 per bin (14 hours × $4.25 = $59.50 compared with 9 bins × $6.50 = $58.50; 13 hours 45 minutes at the minimum wage would yield $58.43, which is less than the piece rate.). Because of the available daylight, that situation could only occur in June and would probably require the picker to forego his 1/2 hour uncompensated lunch break. If there were a picker with such stamina to work for 14 hours with little or no meal break, he (or she) would most likely be a lion (or lioness) who would pick more than nine bins in a 14-hour day.

Matching at 10 bins would require more than a seemingly impossible 15-hour workday in order to affect a picker's piece rate compensation (15 hours × $4.25 = $63.75 compared with 10 bins × $6.50 = $65.00).

[5] In this testimony, Rollins appears to use the word "match" differently than matching hours to bins. Rather, he is indicating that Dieujuste was not matching his records to the actual events.

[6] The January 7, 1993, memorandum relied upon by the plaintiffs did not purport to state the details of the incentive program; it was simply a brief explanation of the first paycheck of the calendar year. The plaintiffs thus failed to adduce probative evidence that the attendance requirements were not a part of the program throughout the time the program was in effect.

[7] The plaintiffs' post-trial submission did assert a claim on behalf of Efren Zepeda. However, as the defendants correctly point out, there was no claim alleged for Zepeda in the Amended Complaint concerning bonus or incentive payments for the 1992-1993 season. Zepeda, therefore, is not entitled to any relief on such a claim. In all events, the pay records show a gap in Zepeda's work of more than three days in March 1993, and thus he would not qualify for the bonus or incentive payments. Significantly, Zepeda testified at the trial, but made no attempt to explain his absences.

[8] It is noted, specifically, that the plaintiffs in their post-trial submissions do not assert that the checks and check stubs used in the 1991-1992 season and the beginning of the following season lacked the required information. Consequently, the plaintiffs have not demonstrated for that time frame a violation for the failure to provide piece rate information.

[9] The plaintiffs do not base their claim that the AWPA violations were intentional (or that the FLSA violations were willful) upon the actions of the crew leaders (Doc. 249, pp. 53-54, 61-62).

[10] The plaintiffs, in connection with a claim of liquidated damages under the FLSA, refer to a defense of good faith. The defendants, however, have not discussed that issue, and, thus, any consideration of that matter will be deferred. Further, the plaintiffs asserted in their reply brief a tolling argument regarding the statute of limitations. That issue also has not been adequately explicated and will therefore be deferred.