Diaz v. Eagle Produce

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT ABEL RUIZ DIAZ; UBALDO MORENO;  PIEDAD H. RENTERIA; ALEJANDRO D. MANCILLA, Plaintiffs-Appellants, No. 06-15878 v.  D.C. No. CV-03-02127-MHM EAGLE PRODUCE LIMITED PARTNERSHIP; PHOENIX AGRO- OPINION INVEST, INC.; SAM MANAGEMENT INC., Defendants-Appellees.  Appeal from the United States District Court for the District of Arizona Mary H. Murguia, District Judge, Presiding Argued and Submitted February 15, 2008—San Francisco, California Filed April 4, 2008 Before: William C. Canby, Jr., and Milan D. Smith, Jr., Circuit Judges, and Stephen G. Larson,* District Judge. Opinion by Judge Milan D. Smith, Jr. *The Honorable Stephen G. Larson, United States District Judge for the Central District of California, sitting by designation. 3519 DIAZ v. EAGLE PRODUCE 3523 COUNSEL George H. McKay, Community Legal Services, Phoenix, Ari- zona, for the plaintiffs-appellants. Emily M. Craiger and J. Greg Coulter, Littler Mendelson, P.C., Phoenix, Arizona, for the defendants-appellees. OPINION MILAN D. SMITH, JR., Circuit Judge: In this appeal we consider whether the district court appro- priately entered summary judgment against four farm workers who brought suit under the Age Discrimination in Employ- ment Act (ADEA), 29 U.S.C. § 621 et seq., after being dis- charged in the context of a seasonal slowdown in agricultural activity. Applying the three-stage burden-shifting framework established in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973), we affirm in part, reverse in part, and remand for trial on one worker’s claim of age discrimination. I. BACKGROUND Defendants-Appellees Phoenix Agro Invest, Inc., and SAM Management, Inc., are general partners in Defendant- 3524 DIAZ v. EAGLE PRODUCE Appellee Eagle Produce Limited Partnership (Eagle Produce), which operates a commercial broccoli and melon farm in Aguila, Arizona. The seasonal nature of farming these crops creates a fluctuating need for laborers that corresponds with periods of soil preparation, crop planting, and harvest. Eagle Produce hires new workers or transfers existing employees to satisfy its labor needs during periods of increased activity. Conversely, it lays off workers or transfers them to other operations when its need for labor decreases. Because work on the farm generally slows down each year between the melon harvest in the fall and planting in the spring, layoffs and transfers tend to occur in the winter. Reflecting this trend, the number of workers employed by Eagle Produce was 770 in July 2002, but only 332 in February of the same year. Plaintiffs-Appellants Abel Ruiz Diaz, Ubaldo Moreno, Pie- dad H. Renteria, and Alejandro D. Mancilla (collectively, Plaintiffs) were laborers at Eagle Produce’s Aguila farm. With the exception of Alejandro Mancilla, they were mem- bers of a tractor crew known as Crew 94 that was responsible for preparing the soil for planting. Plaintiffs were supervised exclusively by Chet Daffern until May 2001, when Eagle Pro- duce hired Owen Brandt to work as co-supervisor. Brandt became the sole supervisor of Crew 94 when Daffern quit at the end of December 2001. One of Brandt’s responsibilities was to make crew personnel decisions. The work available at the farm became particularly scarce in the winter of 2001-2002. This was at least partly due to the farm’s increased use of plastic mulching, a technique that involves placing a thin layer of plastic over the soil to protect it from the effects of the local climate. Fields covered with plastic mulch, as opposed to other forms of mulch, require significantly fewer hours to prepare. Even less preparation is required for fields that use so-called second-year plastic mulch, or plastic mulch laid down the previous year. From 2001 to 2002, the use of plastic mulch on Eagle Produce’s cantaloupe fields increased from 2191.88 acres to 2690.25 DIAZ v. EAGLE PRODUCE 3525 acres. Within those totals, the acreage of second-year plastic mulch increased from 287.4 to 1208.3 acres. The use of plas- tic mulch for Eagle Produce’s honeydew and mixed melon crops similarly increased from 70.4 acres in 2001 to 186.3 acres in 2002. The efficiency gained through the expanded use of this technique allowed Crew 94 to prepare the same amount of land with fewer hours of labor. The improved effi- ciency in turn reduced Crew 94’s combined total work hours from 13,710.75 hours in January and February 2001 to only 7,346 for the same two-month period in 2002. Brandt laid off several workers during this period. He claims that in deciding whose employment to terminate, he did not consider the extent of the employee’s experience with a tractor, how many years the employee had been working at Eagle Produce, or the employee’s age or wage. Brandt did, however, consider job performance, attitude, attendance, work ethic, and the individual’s ability to work with others. Plain- tiffs were among those whose employment was terminated. The circumstances pertaining to each plaintiff are as follows: Abel Ruiz Diaz Diaz began working for Eagle Produce in August 1997 at the age of 51. At the time, he had approximately seven years of farm work experience. He drove a tractor at Aguila and, during harvest season, drove a truck from which he sprayed water onto the ground to reduce dust problems. Diaz was not qualified to drive a planting tractor, a responsibility that requires a certain level of technical expertise. Other than missing work approximately once a month due to a drinking problem, Diaz adequately fulfilled his duties at the farm. On January 19, 2002, Brandt informed Diaz that he no lon- ger had any work available. He offered Diaz a temporary transfer to the harvest tractor crew, but Diaz declined the offer because he was unfamiliar with that job, it was temporary, it involved fewer hours, and, in his view, it would be difficult 3526 DIAZ v. EAGLE PRODUCE to perform in light of a medical condition that precluded him from lifting objects heavier than fifty pounds. Shortly thereaf- ter, Diaz was laid off. He was 55 years old at the time. An Eagle Produce notice states that he was laid off “due to reduc- tion of work.” Ubaldo Moreno Moreno began working for Crew 94 as a tractor and water- truck driver in June 2000, at the age of 65. He was a depend- able worker, but had a history of damaging Eagle Produce property, including an irrigation ditch in June 1998 and a “disc and tractor” in March 2000. He also allowed unautho- rized persons to drive Eagle Produce equipment in October 2001. On February 5, 2002, a tractor broke down while he was driving it, necessitating approximately $10,000 in repairs. Eagle Produce laid off Moreno the next day due to “reduction of work.” He was 66 years old. Moreno’s “history of damag- ing” Eagle Produce property was considered in selecting him as one of the workers to lose his job. Piedad H. Renteria Renteria worked as a tractor driver for Crew 94 and as a planting-tractor driver.1 He also operated his own check- cashing business at the farm in violation of a company policy that prohibited solicitation on Eagle Produce property. On paydays, he would park his truck in front of the farm’s office and cash his coworkers’ paychecks in exchange for a small fee. He maintained this business for several years without ever being reprimanded or told to stop. Once Brandt became aware of the business, he warned Renteria to stop. Brandt gave the warning orally with the assistance of a translator to ensure that Renteria understood. Renteria indicated that he 1 It is unclear whether he began this employment in 1978 at the age of approximately 41 or in 1987 at the age of 51, but the difference is ulti- mately inconsequential. See infra § III.A.1. DIAZ v. EAGLE PRODUCE 3527 understood the warning, but nevertheless promptly resumed his business. Brandt dismissed him on February 23, 2002. Renteria was 65 years old at the time. Eagle Produce explains that Renteria was “laid off due to a seasonal work slow down, reduction in work and repeated violations of rules established in the Company Handbook.” Renteria was not the only employee who cashed checks on Eagle Produce property, but, to his knowledge, he was the only one disciplined for doing so. Alejandro Mancilla Eagle Produce hired Mancilla when he was 58 in October 1996 to test GPS equipment on farm tractors. After that task was completed, Brandt hired Mancilla to work as a water- truck driver. In that capacity, Mancilla accidentally broke the bottom of a chemical trailer on January 23, 2002, and once received a verbal warning for failure to wear safety equip- ment. As work became unavailable at the end of the melon harvest, Brandt laid him off on February 5, 2002. Mancilla was 63 years old at the time. The employment-action notice states that he was laid off “due to reduction of work,” but Eagle Produce also considered the damage that Mancilla caused to the trailer and his failure to wear safety equipment in deciding to terminate his employment. Mancilla never worked for Crew 94 when Brandt was the supervisor and was not qualified for planting. Eagle Produce has indicated that Mancilla is eligible for rehire when more work becomes available. Plaintiffs filed an action in district court seeking declara- tory relief and damages on the theory that their terminations violated the ADEA, the Migrant and Seasonal Agricultural Worker Protection Act, 29 U.S.C. § 1854 et seq., and the Americans with Disabilities Act, 42 U.S.C. § 12111 et seq. The district court granted Defendants-Appellees’ motion for summary judgment on all claims. The workers appeal only the disposition of their ADEA claim. 3528 DIAZ v. EAGLE PRODUCE II. STANDARD OF REVIEW AND JURISDICTION We review de novo a district court’s grant of summary judgment pursuant to Federal Rule of Civil Procedure 56. Buono v. Norton, 371 F.3d 543, 545 (9th Cir. 2004). Rule 56(c) provides that summary judgment is warranted when the “pleadings, the discovery and disclosure materials on file, and any affidavits show that there is no genuine issue as to any material fact and that the movant is entitled to judgment as a matter of law.” A “genuine issue” of material fact will be absent if, upon “viewing the evidence and inferences which may be drawn therefrom in the light most favorable to the adverse party, the movant is clearly entitled to prevail as a matter of law.” Jones v. Halekulani Hotel, Inc., 557 F.2d 1308, 1310 (9th Cir. 1977). Summary judgment is inappropri- ate if reasonable jurors, drawing all inferences in favor of the nonmoving party, could return a verdict in the nonmoving party’s favor. United States v. Shumway, 199 F.3d 1093, 1103-04 (9th Cir. 1999). We have jurisdiction pursuant to 28 U.S.C. §§ 1331 and 1291. III. DISCUSSION [1] The ADEA makes it unlawful “to discharge any indi- vidual . . . because of such individual’s age.” 29 U.S.C. § 623(a)(1). We evaluate ADEA claims that are based on cir- cumstantial evidence of discrimination by using the three- stage burden-shifting framework laid out in McDonnell Doug- las Corp. v. Green, 411 U.S. 792 (1973). Enlow v. Salem- Keizer Yellow Cab Co., 389 F.3d 802, 812 (9th Cir. 2004). Under this framework, the employee must first establish a prima facie case of age discrimination. Coleman v. Quaker Oats Co., 232 F.3d 1271, 1281 (9th Cir. 2000). If the employee has justified a presumption of discrimination, the burden shifts to the employer to articulate a legitimate, non- discriminatory reason for its adverse employment action. Id. DIAZ v. EAGLE PRODUCE 3529 If the employer satisfies its burden, the employee must then prove that the reason advanced by the employer constitutes mere pretext for unlawful discrimination. Id. “As a general matter, the plaintiff in an employment discrimination action need produce very little evidence in order to overcome an employer’s motion for summary judgment.” Chuang v. Univ. of Cal. Davis, Bd. of Trs., 225 F.3d 1115, 1124 (9th Cir. 2000). A. The Prima Facie Cases [2] Each plaintiff can establish a prima facie case of dispa- rate treatment by demonstrating that he was (1) at least forty years old, (2) performing his job satisfactorily, (3) discharged, and (4) either replaced by substantially younger employees with equal or inferior qualifications or discharged under cir- cumstances otherwise “giving rise to an inference of age dis- crimination.” Coleman, 232 F.3d at 1281.2 An inference of discrimination can be established by “showing the employer had a continuing need for [the employees’] skills and services in that their various duties were still being performed . . . or by showing that others not in their protected class were treated more favorably.” Id. (quotation marks and citation omitted). 2 Generally, an employee can satisfy the last element of the prima facie case only by providing evidence that he or she was replaced by a substan- tially younger employee with equal or inferior qualifications. Id. The test for the prima facie case changes somewhat, however, where a discharge occurs in the context of a general reduction in the employer’s workforce. In this context, circumstantial evidence other than evidence concerning the identity of a replacement employee may also warrant an inference of dis- crimination. Id.; Ritter v. Hughes Aircraft Co., 58 F.3d 454, 457 (9th Cir. 1995). The reason for this difference is that in most reduction-in-force cases no replacements will have been hired. We presently apply the modi- fied test because there is no genuine dispute that the discharge of Plaintiffs occurred in the context of a general reduction in the number of workers employed at the Aguila farm in the winter of 2001-2002. 3530 DIAZ v. EAGLE PRODUCE [3] Plaintiffs were each over forty years old and were dis- charged after the commencement of their protected status. Eagle Produce does not dispute that these elements of the workers’ prima facie cases have been satisfied. The only questions, then, are whether Plaintiffs were performing their jobs satisfactorily and whether they were discharged under circumstances giving rise to an inference of age discrimina- tion. For the reasons explained below, we conclude that the dis- trict court appropriately granted summary judgment against Renteria because he failed to create a triable issue concerning whether his job performance was satisfactory. We also con- clude, however, that the district court erred in granting sum- mary judgment against Diaz, Moreno, and Mancilla based on the first stage of the McDonnell Douglas analysis. Those plaintiffs provided evidence of satisfactory job performance and an accumulation of circumstantial evidence that, taken together, could lead reasonable jurors to draw an inference of age discrimination. 1. Whether the workers’ job performances were satisfactory [4] Diaz, Mancilla, and Moreno have each established a tri- able issue of fact regarding the second element of the prima facie case. Diaz and Mancilla generally performed depend- ably and without incident. Deficiencies in their performance were relatively minor and infrequent. The issue is closer for Moreno because he damaged Eagle Produce property three times over the course of approximately four years and once violated a company safety rule, but those incidents were also relatively infrequent. Moreover, there is no evidence that Moreno caused any of the damage intentionally or recklessly, and his supervisor generally found him to be dependable. [5] The same cannot be said of Renteria, who over an extended period openly violated Eagle Produce’s policy DIAZ v. EAGLE PRODUCE 3531 against solicitation on company property and continued to do so even after receiving a warning from Brandt. Considering that behavior, no reasonable juror could find that Renteria’s performance was satisfactory. See Mungro v. Giant Food, Inc., 187 F. Supp. 2d 518, 522 (D. Md. 2002) (citing cases for the proposition that “[a] plaintiff who violates company pol- icy and fails to improve his performance despite a warning has not demonstrated satisfactory performance”). Because Renteria has failed to establish a prima facie case of discrimi- nation, we do not complete the remainder of the McDonnell Douglas analysis with respect to his claim. [6]2. Whether the circumstances give rise to an inference of discrimination Although it is a close question, we conclude that Diaz, Mancilla, and Moreno have also created a triable issue of fact concerning the fourth and final requirement for a prima facie case. This conclusion draws support from a variety of circum- stantial evidence. i. Statistical evidence Including Plaintiffs, the average age of the sixteen workers laid off or discharged from Crew 94 in 2001 and 2002 is 48.4.3 3 We derive the figures used in this analysis from the employees’ dates of birth, hire, and last pay check, as listed in the Eagle Produce Employees Chart and amended by the exhibits attached to Defendants-Appellees’ Supplemental Statement of Facts in Support of Their Motion for Summary Judgment. At oral argument, Eagle Produce suggested that the Chart was an improper basis for a statistical analysis. If the argument is that the Chart is inadmissible under the Federal Rules of Evidence, we reject the argu- ment as waived. The district court already ruled that the Chart is admissi- ble, and Eagle Produce did not contest that decision in its brief on appeal. See Greenwood v. FAA, 28 F.3d 971, 977 (9th Cir. 1994) (“We review only issues which are argued specifically and distinctly in a party’s open- ing brief.”). If Eagle Produce is arguing simply that the Chart contains inaccuracies, our analysis accounts for them by incorporating into the Chart the amendments provided in Defendants-Appellees’ Supplemental Statement of Facts. 3532 DIAZ v. EAGLE PRODUCE Three of those workers were 20, 21 and 30 years old, respec- tively. While the majority fell within the age category pro- tected by the ADEA, most were also hired after the age of forty. One was hired at 69 years of age, only a year and a half before the disputed layoffs. Another was hired in July 2001 at age 65. Several others were hired when they were in their mid- to late-forties. Sixteen other workers were hired for Crew 94 in 2001 and 2002 and not subsequently discharged. Their average age at the time of hiring was 38.75. Nearly half of them were 40 or more years old, including one who was 63 and two who were 57. On its own, this evidence would suggest that an inference of discrimination is not warranted. Eagle Produce laid off many of the Crew 94 employees, including Plaintiffs, only shortly after hiring them at what were already relatively advanced ages. Diaz was hired at 51 and laid off four years later at 55. Moreno was hired at 65 and laid off at 66. Man- cilla was hired at 58 and laid off at 63. If Eagle Produce was biased against older workers, it presumably would not have hired Plaintiffs in the first place. The temporal proximity between each Plaintiff’s hiring and layoff also makes it unlikely that age later developed as the reason for the dis- charges. The difference in physical and mental capacity between an average 65 year-old and an average 66 year-old, or between a 58 year-old and a 63 year-old, is not significant enough to warrant an inference of anything but the most arbi- trary bias. See Bradley v. Harcourt, Brace & Co., 104 F.3d 267, 270-71 (9th Cir. 1996) (“[W]here the same actor is responsible for both the hiring and the firing of a discrimina- tion plaintiff, and both actions occur within a short period of time, a strong inference arises that there was no discrimina- tory motive.”). That the average age of the workers hired during the two- year period is approximately nine and a half years younger DIAZ v. EAGLE PRODUCE 3533 than the average age of those laid off—38.75 versus 48.4 years—also fails to justify an inference of age discrimination. The disparity is not so stark as to suggest bias rather than pure chance. See O’Connor v. Consol. Coin Caterers Corp., 517 U.S. 308, 313 (1996) (“In the age discrimination context . . . an inference [of discriminatory intent] cannot be drawn from the replacement of one worker with another worker insignifi- cantly younger.”); Hartley v. Wis. Bell, Inc., 124 F.3d 887, 893 (7th Cir. 1997) (holding that replacement employees must be at least ten years younger than their predecessors to justify a presumption of age discrimination). This is particularly true because the two data sets of sixteen workers are too small to form a reliable basis for analysis. See Palmer, 794 F.2d at 539 (finding certain statistical evidence of employment discrimi- nation unpersuasive because the sample size was too small). [7] A different picture emerges, however, when we con- sider the data with Brandt in mind. He first began to make personnel decisions for Crew 94 when he was hired as a supervisor in May 2001. The average age of the workers hired before that date was 44.29. For the period of May 2001 to January 2002, during which Daffern and Brandt both made personnel decisions, the average age of Crew 94 hirees dropped to 40.8. Once Brandt took over as the sole hiring authority, the average age dropped still further to 35.28. By contrast, the average age of workers laid off from Crew 94 increased slightly from 46.2 during the period of Daffern’s and Brandt’s joint supervision to 51.1 after Brandt became the sole supervisor. In short, the disparity between the average age of those hired and those laid off increased from slightly less than two years to nearly 16 years once Brandt started to make personnel decisions. This evidence suggests that although Eagle Produce was not responsible for discrimina- tory hiring practices prior to Brandt’s advent, Brandt used his authority to replace older workers with younger counterparts. [8] Reasonable jurors could find that this interpretation of the data supports an inference of discrimination. Viewing the 3534 DIAZ v. EAGLE PRODUCE statistical evidence with Brandt in mind helps to explain how Eagle Produce could both hire Plaintiffs without regard to age and also terminate their employment because of age shortly thereafter. Because Brandt did not work at Eagle Produce until May 2001, he could not preclude the hiring of Mancilla in approximately 1996, Diaz in 1997, and Moreno in 2000. However, he could lay off these workers because of their ages in the winter of 2002. ii. Knowledge of the workers’ ages [9] The inference of discrimination also draws support from Brandt’s knowledge that his hirees were on average substan- tially younger than the workers he laid off. Brandt did not know the precise ages of Diaz, Moreno, and Mancilla, but he must have possessed a general sense that these individuals were relatively advanced in age because he personally observed them at the farm on a regular basis. See Woodman v. WWOR-TV, Inc., 411 F.3d 69, 80 (2d Cir. 2005) (on-the-job contact is sufficient to warrant an inference of an employer’s knowledge of age). Because of his final position as the sole hiring authority for Crew 94, the crew’s moderate size, and his daily contact with its workers, Brandt also must have known that the individuals he hired in 2002 were younger than those he laid off. Id. In concluding otherwise, the district court failed to construe the evidence in the light most favor- able to Plaintiffs. See Jones, 557 F.2d at 1310. iii. The decision not to fire several younger, less experienced workers [10] Reasonable jurors could find that Brandt’s decision not to lay off several substantially younger workers with less experience than Diaz, Moreno, and Mancilla also supports an inference of discrimination. One worker was 19 years old and had approximately five months of experience in Crew 94 at the time Plaintiffs lost their jobs. Another was 21 years old and had approximately two years of experience in Crew 94. DIAZ v. EAGLE PRODUCE 3535 Another was 36 and possessed about 1.5 years of experience. Still another was 42 and had two months of experience. At the very least, these individuals had no more experience at the Aguila farm than Diaz, Moreno, and Mancilla, who worked there for 4, 1.5, and 5 years, respectively. Nevertheless, Brandt chose not to fire any of these younger workers during the seasonal slowdown. iv. The qualifications of Plaintiffs relative to their replacements [11] Evidence concerning the job experience of Diaz and Moreno relative to that of their replacements provides addi- tional support for an inference of discrimination. Seven indi- viduals were hired around or shortly after the dates on which Plaintiffs lost their jobs. Due to a lack of evidence, the qualifi- cations of three of those hired cannot be determined. Two oth- ers were hired successively to operate a grading machine with which Diaz, Moreno, and Mancilla had no experience. Accordingly, they cannot be viewed as replacements. The remaining two workers were Javier Gastelum (28) and Jesus F. Valdez (21). Gastelum purportedly had 15 years of experi- ence driving a tractor at the time he was hired. Valdez had around five years of experience with a tractor and “knew a lit- tle” about that line of work. This evidence creates a triable issue of fact as to whether Diaz and Moreno were at least as qualified as their succes- sors. Diaz had driven tractors and performed general farm labor for eleven years. Valdez was no more qualified in terms of length of experience. Gastelum had driven a tractor longer than either Diaz or Moreno, but it is unlikely that this differ- ence made Gastelum more qualified in any meaningful sense. The parties do not dispute that the job did not require special- ized training, and the learning curve was not such that years of experience were required for mastery. 3536 DIAZ v. EAGLE PRODUCE [12] The district court erred in weighing against Plaintiffs the absence of evidence that particular, less-qualified, youn- ger workers were hired in their steads. It is true that, due to high turnover in Crew 94, Plaintiffs are unable to identify their individual replacements. However, we treat the last ele- ment of the prima facie case with “flexibility.” Nidds v. Schindler Elevator Corp., 113 F.3d 912, 917 (9th Cir. 1996). To support an inference of discrimination an employee need not demonstrate that one particular individual was designated as his replacement; evidence that a group of younger and comparably or less-qualified employees assumed the plain- tiff’s responsibilities is sufficient. See Rodriguez-Torres v. Carribean Forms Mfr., Inc., 399 F.3d 52, 59 (1st Cir. 2005). It is therefore enough that Diaz and Moreno identified Gas- telum and Valdez as their collective replacements.4 B. Facial legitimacy of the explanations for the layoffs [13] Moving to the next stage of the analysis, we conclude that Eagle Produce did not offer a legitimate, non- discriminatory explanation for the layoff of Diaz. To suffice under McDonnell Douglas, an employer’s explanation must explain why the plaintiff “in particular” was laid off. Davis v. Team Electric Co., ___ F.3d ___, No. 05-35877, slip op. at 3191 (9th Cir. Mar. 28, 2008); see also McDonnell Douglas 4 Unlike Diaz and Moreno, Mancilla has failed to create a triable issue concerning whether he was replaced by younger workers who were equally or less qualified. There is no evidence that he was actually replaced by any of the younger workers. Because Mancilla was not employed in Crew 94, none of the individuals subsequently hired to that group can fairly be viewed as his replacements. Nevertheless, the remain- ing circumstantial evidence is still cumulatively sufficient to permit rea- sonable jurors to infer that Mancilla was discriminated against because of his age. See Wallis v. J.R. Simplot Co., 26 F.3d 885, 889 (9th Cir. 1994) (“The requisite degree of proof necessary to establish a prima facie case for . . . ADEA claims on summary judgment is minimal and does not even need to rise to the level of a preponderance of the evidence.”); Coleman, 232 F.3d at 1281 (evidence of replacement is not essential where a dis- charge occurs in the context of a general workforce reduction). DIAZ v. EAGLE PRODUCE 3537 Corp., 411 U.S. at 803 (concluding that the employer- petitioner satisfied its burden at this stage by specifying “[the employee’s] participation in unlawful conduct . . . as the cause for his rejection”). On its own, the explanation that Diaz was discharged as part of a general reduction in force fails this requirement. Workforce reduction explains why Eagle Produce laid off a group of its workers, but it does not explain why Diaz was chosen to be part of that group. Because no other explanation was given with respect to Diaz, Eagle Pro- duce failed to satisfy its burden at stage two on his claim, and summary judgment was inappropriate.5 [14] Eagle Produce did, however, provide legitimate, non- discriminatory explanations for the layoffs of Moreno and Mancilla by stating that they were each chosen to be part of the reduction in force because of damage caused to company property. Those explanations were individualized, unlike the explanation given to Diaz, and reflected concerns about job performance unrelated to the workers’ protected status. See Pottenger v. Potlatch Corp., 329 F.3d 740, 746 (9th Cir. 2003) (performance-related concerns satisfied stage two under 5 Our prior cases do not require otherwise. Aragon v. Republic Silver State Disposal, Inc., 292 F.3d 654, 661 (9th Cir. 2002), stated in dictum that a seasonal reduction in force and an employee’s poor job performance “[b]oth constitute a legitimate, nondiscriminatory reason for terminating” an employee. Because the employee in Aragon was discharged in the con- text of a reduction in force and selected to be part of the reduction because of poor performance, id. at 657-58, Aragon did not decide whether a reduction in force, standing alone, can suffice under stage two of McDon- nell Douglas. Moreover, the quoted text from Aragon cited exclusively to dictum from the dissenting judge in Winarto v. Toshiba America Electron- ics Componenents, Inc., 274 F.3d 1276, 1295 (9th Cir. 2001) (Wardlaw, J., concurring in part and dissenting in part), another case involving an employee laid off in the context of a reduction in force because of perfor- mance. None of the cases on which the dictum in Winarto relied involved an employer who explained a layoff solely as a product of workforce reduction. See id. (citing Sengupta v. Morrison-Knudsen Co., Inc., 804 F.2d 1072, 1075 (9th Cir. 1986) (citing Gay v. Waiters’ & Dairy Lunch- men’s Union, Local No. 30, 694 F.2d 531, 547 (9th Cir. 1982))). 3538 DIAZ v. EAGLE PRODUCE the McDonnell Douglas framework). The presumption of dis- crimination that was created by the prima facie cases accord- ingly disappears with respect to Moreno and Mancilla. Wallis, 26 F.3d at 892. C. Pretext The final stage of the McDonnell Douglas analysis requires Moreno and Mancilla to raise a genuine issue of fact concern- ing whether the facially legitimate reasons proffered by Eagle Produce are pretextual. Coleman, 232 F.3d at 1281. The workers may demonstrate pretext “either directly by persuad- ing the court that a discriminatory reason likely motivated [Eagle Produce] or indirectly by showing that [the compa- ny’s] proffered explanation is unworthy of credence.” Snead v. Metro. Prop. & Cas. Ins. Co., 237 F.3d 1080, 1093-94 (9th Cir. 2001) (internal quotation marks and citation omitted). Plaintiffs argue that the explanations for their layoffs lack credibility for three reasons. First, they contend that a sea- sonal slowdown did not reduce Eagle Produce’s need for trac- tor workers. Second, they argue that Eagle Produce offered inconsistent explanations for the layoffs. Finally, they contend that the explanations lack credibility because Brandt violated the company handbook by terminating workers without con- sidering the lengths of their employment. We address these arguments in turn. 1. The seasonal slowdown [15] There is no dispute that Eagle Produce’s need for trac- tor workers decreased in the winter of 2001-2002. The total number of employees at the Aguila farm in February 2002, including non-tractor workers, was less than half of the num- ber employed in July of the same year. According to John Redmond, the vice president of Phoenix Agro Invest, the effi- ciency gained by the increased use of plastic mulch reduced Crew 94’s total work hours from 13,710.75 in January and DIAZ v. EAGLE PRODUCE 3539 February 2001 to only 7346 for the same period in 2002. Daf- fern corroborated this testimony. Brandt continued to hire new workers during the slow- down, but that does not suggest that the slowdown did not occur. He hired two workers for Crew 94 in December 2001 and four more in the first two months of 2002, but the evi- dence clearly suggests that the hires were necessitated by the decisions of five other workers to quit during the same period. There is no evidence that Crew 94’s operations ceased alto- gether in the winter. Brandt needed to make some hires to sus- tain a minimum operational level even during the slowdown. Evidence that Eagle Produce’s need for tractor drivers does not decline during the winter similarly fails to create a genu- ine issue of fact. Daffern testified that tractor drivers perform a “consistent” amount of work “year-round.” Gilberto Vigueria testified that there is “more work for tractor drivers from January to June than from July to December.” This is consistent with other evidence that planting for the melon crop begins in the spring. Because Crew 94’s primary duty was to prepare the soil prior to planting, its work for the melon crop would presumably have to begin in the winter, or at least by the early spring. However, none of this evidence actually conflicts with the evidence that the winter of 2002 was particularly slow. The testimony of Daffern and Vigueria addressed the amount of work generally available for the trac- tor crew within a single year; it did not compare the amount of work available in the winter of 2002 to the amount avail- able in previous winters. Thus, even accepting the testimony as true, there is no reason to discount the specific evidence that factors such as the expanded use of plastic mulch caused a unique drop in Eagle Produce’s need for laborers around the time Moreno and Mancilla were laid off.6 6 There is a genuine issue of fact as to whether the need for laborers also declined as a result of Eagle Produce’s decision not to produce a broccoli crop in 2002. Brandt testified to that effect, but company records and Ste- 3540 DIAZ v. EAGLE PRODUCE [16] From the conclusion that a seasonal slowdown occurred, however, it does not necessarily follow that there is no triable issue of fact concerning pretext. The slowdown could justify only a workforce reduction that comparably affects laborers from different age groups. It cannot explain the statistical evidence suggesting that Brandt had a tendency to fire older workers and hire substantially younger replace- ments. It also cannot explain the evidence that Brandt fired Moreno and Mancilla while retaining younger workers with less experience and later hired younger workers who were for the most part no more qualified than Plaintiffs. We therefore conclude that, even if the slowdown-induced reduction in force could independently stand as a sufficiently individual- ized explanation for the layoffs under stage two of McDonnell Douglas, there would be a triable issue of fact concerning pre- text to the extent that the explanation was unaccompanied by any employee-specific justification for discharge. 2. The consistency of Eagle Produce’s explanations for the layoffs We reject the premise of Plaintiffs’ argument that inconsis- tency between the explanations for the layoffs renders the explanations not credible. Eagle Produce proffered at least two justifications for the layoffs of both Moreno and Mancilla —an overall workforce reduction and performance deficien- cies. Those explanations were each different, but not inconsis- tent. The reasonable interpretation is that the seasonal downturn necessitated a certain number of layoffs and that Eagle Produce considered job performance in deciding whom to lay off. See Aragon, 292 F.3d at 661 (“We do not infer pre- phen A. Martori, one of Eagle Produce’s limited partners, contradict him. Construing this evidence in the light most favorable to Plaintiffs, reason- able jurors could conclude that the work available for Crew 94 did not appreciably decline in the winter as a result of a decision to not produce broccoli. However, that conclusion does not undermine the other evidence that work became scarce because of the expanded use of plastic mulch. DIAZ v. EAGLE PRODUCE 3541 text from the simple fact that [an employer] had two different, although consistent, reasons for laying off” an employee.). 3. Brandt’s failure to follow the company handbook [17] Plaintiffs also argue that summary judgment is inap- propriate in part because of evidence that Brandt violated a company policy in terminating their employment. We agree. Brandt testified that he did not consider the length of the workers’ employment at the Aguila farm prior to laying them off. This violated Eagle Produce’s company handbook, which required him to consider “skill, ability, attendance, production records, and the length of employment.” Reasonable jurors could conclude that this irregularity further undermines the credibility of the proffered explanations for the layoffs: if age was truly irrelevant to Brandt’s decisionmaking, he presum- ably would not have failed to weigh the factor in the hand- book that weighed most heavily in favor of retaining older workers. The evidence is consistent with the view that Brandt disregarded company policy because it conflicted with his intent to discriminate. See Brennan v. GTE Govt. Sys. Corp., 150 F.3d 21, 29 (1st Cir. 1998) (“Deviation from established policy or practice may be evidence of pretext.”). 4. Moreno and Mancilla have not presented sufficient evidence to avoid summary judgment on pretext [18] Notwithstanding the foregoing analysis, we find that Moreno has failed to create a genuine issue of fact concerning pretext. The evidence that he caused significant damage to Eagle Produce property over the course of several years is undisputed. Brandt testified that he did not fire Moreno on the basis of two incidents of property damage that occurred before May 2001, but Moreno was also responsible for other damage in October 2001 and February 2002. Moreno fails to account for evidence that these latter incidents were the genu- 3542 DIAZ v. EAGLE PRODUCE ine basis for his termination. The February 2002 incident alone resulted in $10,000 in repairs. That Brandt fired Moreno the day after that incident leaves little doubt that the property damage, rather than age, motivated Brandt’s decision.7 [19] Mancilla has similarly failed to establish a genuine issue concerning the credibility of the explanation for his ter- mination. Brandt testified that he laid off Mancilla because there was no need for water truck drivers in the winter after the close of the melon harvest. That testimony is uncontro- verted and establishes that seasonal slowdown was in fact the reason for Mancilla’s discharge. Our conclusions that the seasonal slowdown cannot inde- pendently explain the indicia of age discrimination in the sta- tistical evidence and that Brandt failed to follow the company handbook do not undermine the credibility of Eagle Produce’s explanations for the layoffs of Mancilla and Moreno. Even with such general circumstantial evidence of discrimination, no reasonable juror could find pretext in light of the specific and uncontroverted evidence that nondiscriminatory consider- ations motivated the individual discharges of these two work- ers. 7 This conclusion is not inconsistent with the finding that Moreno pro- vided sufficient evidence of satisfactory job performance to avoid sum- mary judgment at the first stage of the McDonnell Douglas analysis. The focus of the pretext inquiry is not to determine whether Eagle Produce was correct in determining that Moreno’s job performance was unsatisfactory, but simply whether Moreno’s performance was the real reason for the ter- mination. Douglas v. Anderson, 656 F.2d 528, 533 n.5 (9th Cir. 1981); see also DeJarnette v. Corning Inc., 133 F.3d 293, 299 (4th Cir. 1998) (“[W]hen an employer articulates a reason for discharging the plaintiff not forbidden by law, it is not our province to decide whether the reason was wise, fair, or even correct, ultimately, so long as it truly was the reason for the plaintiff’s termination.”) (internal quotation marks and citation omitted). DIAZ v. EAGLE PRODUCE 3543 IV. CONCLUSION For the foregoing reasons, we hold that Renteria failed to create a triable issue of fact as to whether his job performance was satisfactory, and thus failed to establish a prima facie case of age discrimination. Summary judgment was appropri- ately entered against him. We also hold, however, that Diaz, Moreno, and Renteria provided evidence of satisfactory job performance and an accumulation of circumstantial evidence that, taken together, could lead a reasonable juror to draw an inference of age discrimination. Eagle Produce did not pro- vide a facially legitimate explanation for Diaz’s discharge in stating that the discharge was part of a reduction in force necessitated by a seasonal slowdown in agricultural activity. Summary judgment was therefore inappropriate with regard to Diaz. Eagle Produce’s explanations for the discharges of Moreno and Mancilla were facially legitimate, but Moreno and Mancilla failed to create a genuine issue as to the credi- bility of those explanations. The district court therefore appro- priately entered summary judgment against Moreno and Mancilla. We AFFIRM the entry of summary judgment on the ADEA claims of Ubaldo Moreno, Piedad H. Renteria, and Alejandro D. Mancilla. We REVERSE the entry of summary judgment on the ADEA claim of Abel Ruiz Diaz and REMAND for trial on that claim. The parties will bear their own costs on appeal.