MEMORANDUM AND ORDER

GERTNER, District Judge.West Page 96

I. INTRODUCTION

Plaintiffs, three African American males and one Spanish-surnamed male, bring individual actions and a class action against the Boston Herald, Inc. ("the Herald"). Plaintiffs applied for entry-level jobs at the Herald and were not hired. They allege that the defendant engaged in a variety of discriminatory practices that prevented them, and others similarly situated, from being hired.

While the plaintiffs bring claims under a succession of related counts, they allege essentially both disparate treatment and disparate impact on the basis of race. Disparate treatment is overt, intentional discrimination; disparate impact involves discrimination through ostensibly neutral rules and practices whose net effect is to discriminate.

The undisputed evidence in this case suggests that both kinds of discrimination went on in the Herald pressroom for some substantial period of time. From the moment they applied, applicants for pressroom jobs who were friends and relatives of existing Herald employees were given preference. They received longer, more detailed application forms than did outsiders, forms that invited them to list names of friends and relatives at the Herald. Only those who had friends and relatives were hired. Since existing employees were almost uniformly white, this procedure provided an opportunity for the Herald to discriminate on the basis of race. Statistics regarding the pressroom workforce suggest that intentionally or not, that opportunity was realized. The pressroom was, until quite recently, almost uniformly white.

The heart of the dispute between the parties is how long these policies continued, and whether they were ever applied to or affected these plaintiffs. The plaintiffs assert that these policies continue to this day. Moreover, even when they ended, their discriminatory effects continued; the defendant hired new employees from the applications kept on file under the nepotistic system. The defendant responds with evidence that the discriminatory policies ended in the spring of 1994, months before the first plaintiff applied.

A number of motions are currently before the Court, including plaintiffs' motion for class certification, defendant's motion for summary judgment, and plaintiffs' motion for partial summary judgment. Evidence suggests that the discriminatory system may well have continued into September 1994; during that month one of the plaintiffs applied to the Herald, and two positions were filled by white men who may have had contacts at the paper. Summary judgment to the defendant is accordingly DENIED IN PART and ALLOWED IN PART. Summary judgment to the plaintiffs is DENIED. A class will be certified as described below.

II. PROCEDURAL STANDARD

A. Motion for Summary Judgment

Summary judgment must be granted if "the pleadings, depositions, answers to interrogatories, and admissions on file, together with affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law." Fed.R.Civ.P. 56(c). A "material fact" is one which, if taken to be true, might affect the outcome of the suit. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). In determining whether a genuine dispute as to any material fact exists, the court must view the record in the light most favorable to the non-moving party, indulging all reasonable inferences in that party's favor. Udo v. Tomes, 54 F.3d 9, 12 (1st Cir. 1995); Griggs-Ryan v. Smith, 904 F.2d 112, 115 (1st Cir. 1990). However, where the moving party makes an initial showing that there exists no genuine issue of material fact, the non-moving party may not rely merely on allegations or unsupported speculations to avoid summary judgment. Byrd v. Ronayne, 61 F.3d 1026, 1030 (1st Cir. 1995); Smith v. Stratus Computer, Inc., 40 F.3d 11, 13 (1st Cir. 1994), cert. denied, 514 U.S. 1108, 115 S.Ct. 1958, 131 L.Ed.2d 850 (1995). It must set forth specific facts showing that there is a genuine issue for trial. Anderson, 477 U.S. at 256. As the parties in this case have both moved for summary judgment, I must review the evidence in the light most favorable to each in turn. I will address the defendant's motion first.

B. Plaintiffs' Motion for Class Certification

The plaintiffs have proposed a class consisting of:

(1) past and present Herald pressroom job applicants;

(2) those who will in future apply for employment in the Herald pressroom;

(3) applicants who seek to apply for Herald pressroom jobs and who have been deterred from applying because of the virtually all white racial composition of the Herald pressroom and its reputed and long-standing discriminatory recruitment, selection, and hiring policies and practices; and,

(4) would-be applicants who are deprived of information concerning Herald pressroom job vacancies and who are not recruited for such vacancies.

Rule 23 sets out four requirements for certification of a class: the class is sufficiently numerous that joinder of all potential class members is impracticable; there exist questions of law or fact common to the class; the named plaintiffs' claims are typical of those of the class; and the named plaintiffs will adequately and fairly represent the interests of the class. Fed.R.Civ.P. 23.

Both the typicality and the numerosity requirements will turn to some extent on the success of the defendant's motion for summary judgment. An evaluation of the typicality of the plaintiffs' claims is closely related to an evaluation of their individual claims. Harris v. White, 479 F. Supp. 996, 1008-09 (Mass. 1979). Plaintiffs must show that they themselves have been injured, not merely that an injury has been suffered by members of a class to which they belong. East Texas Motor Freight System, Inc. v. Rodriguez, 431 U.S. 395, 403-404, 97 S.Ct. 1891, 52 L.Ed.2d 453 (1977); Warth v. Seldin, 422 U.S. 490, 502, 95 S.Ct. 2197, 45 L.Ed.2d 343 (1975). If there are no material issues of fact with respect to the plaintiffs' claims and the defendant is entitled to judgment as a matter of law, then the plaintiffs' individual claims are dismissed. A fortiori, they cannot represent the class.

Likewise, the numerosity of the potential class of plaintiffs cannot be determined without first assessing how many of the plaintiffs' claims have merit, and for what period of time. Thus, it is clear that our initial inquiry must involve an evaluation of plaintiffs' individual claims. The first step, then, is to consider the motions for summary judgment, evaluating the defendant's and the plaintiffs' motions in turn.

III. BACKGROUND

The genesis of this case lies in a previous one. In Grant v. News Group Boston, 55 F.3d 1 (1st Cir. 1995), a terminated black employee, Otis Grant ("Grant"), sued the Herald, alleging that its decision to terminate his employment was motivated by racial discrimination. The trial court entered summary judgment for the defendant and the plaintiff appealed. On appeal, the First Circuit upheld the district court, finding that the Herald had shown legitimate, non-discriminatory reasons for Grant's termination. Specifically, the Herald demonstrated that Grant frequently failed to cover his assigned shifts, and that white substitute paperhandlers with equally poor records were also fired. Id. at 3.

The First Circuit also upheld the district court's foreclosure, on procedural grounds, of Grant's challenges to pressroom hiring and application procedures. However, the court expressed serious concern regarding these procedures:

Grant's statistical evidence does paint a disturbing picture of the Herald's pressroom hiring practices and their possible effects. It is apparent that qualified African Americans are significantly under-represented in the Herald's pressroom. Moreover, Robert Reilly [the pressroom supervisor] concedes that word-of-mouth communication and nepotism play a large role in determining who learns about and obtains available paperhandler positions. Finally, in response to an inquiry posed at his deposition, Reilly, who enjoys nearly unfettered discretion over pressroom hiring, expressed little or no concern about the exclusionary effect these facially-neutral practices might be having on potential applicants of color. Grant, 55 F.3d at 8.

This litigation takes up the issues left unresolved in the Grant case, directly challenging the Herald's processes for hiring substitute and full-time paperhandlers in its pressroom.

IV. FACTS

A. The Job of a Paperhandler

Herald paperhandlers prepare paper for the presses. They move rolls of paper, clean waste and ink from the floors, and bale waste paper. It is essentially an entry-level job, for which no specific experience or training is required. Deposition of Robert Reilly, former pressroom superintendent, at 95 (hereinafter "Reilly Dep."). The Herald employs both part-time (also known as substitute) and full-time paperhandlers. Substitute paperhandlers are generally called to work on short notice, during busy periods or to cover for absent full-timers. All of the Herald's full-time paperhandlers have been promoted from substitute paperhandler positions. Id. at 59, 109.

B. The Nepotistic Policies in Place from 1989-1994

From 1989 until July 1996, all pressroom hiring decisions were made by the pressroom superintendent, Robert ("Bubba") Reilly ("Reilly"), with the help of the assistant superintendent, Daniel Messing ("Messing"). Messing has described the hiring system in place until at least 1994 as "to some extent" nepotistic. Deposition of Daniel Messing at 16 (hereinafter "Messing Dep."). In 1994, Reilly admitted that the only way to hear about vacancies at the paper was through a present employee. Reilly Dep. at 60, 62. Typically, after having heard about openings from friends or relatives at the Herald, applicants would go to see Reilly directly to obtain an application form. Id. at 92. They would fill out a yellow, two-sided application form that asked for names of friends or relatives already employed at the Herald, as well as for military or employment history. By contrast, applicants without connections at the Herald would not be sent to see Reilly directly. They would receive a white, one-sided application form at the security desk, a form that did not ask for names of friends or relatives or employment history. Id. at 94. Until at least April 1994, only those who had filled out two-sided applications were considered for pressroom positions. Id. at 24, 101.

During this time — from 1989 to 1994 — the Herald's overall workforce was almost entirely white. EEOC reports filed in this case show that in 1992, the Herald had 865 employees, of whom 27 were black and 5 Hispanic. In 1993, the respective numbers were 842, 30, and 4. In 1994, when the first plaintiff applied to the pressroom, out of 851 employees, 30 were black and 4 were Hispanic.

Not surprisingly, preference to friends and relatives of this workforce resulted in an almost entirely white pressroom. There was not a single black pressroom employee from 1948, when Reilly began work at the Herald's predecessor, the Record American, until the late 1980s — except for one man who worked a single Saturday night at some point during the 1970s. Between December 1989, when Reilly became pressroom superintendent, until April 1996, Reilly hired no black or Spanish-surnamed full-time paper handlers, no Spanish-surnamed substitute paperhandlers, and only one black substitute. Reilly Dep. at 53. Yet twenty-four substitute paperhandlers were hired during that same time period. Reilly Dep. at 26. When asked at his Grant case deposition in February 1994 why there was only one black employee in the pressroom, Messing responded that it was probably because most of the employees had been recommended from within. Messing. Dep. at 16.

C. When Did the Nepotism Policy End?

The Herald states, however, that this word-of mouth system ended some time in 1994, around the time of the Grant suit. Messing Dep. at 27. It maintains that its current policy is to give the one-sided form to all applicants and file the applications according to the date they were submitted. When there is a vacancy in the pressroom, the Herald calls the applicants on file in the order in which the applications were received. It offers the job to the first applicant reached by phone. Although it does not send out written notices of any kind, or recruit minority applicants, id. at 32-33, the Herald maintains that its hiring system is now neutral in intent and effect.

When the word-of-mouth system ended is central to this dispute. It determines whether the plaintiffs themselves were ever affected by it. Plaintiff David Gaines ("Gaines") submitted an application for a paperhandler position to the Herald on September 1, 1994; he was not hired. Plaintiffs Frank Poindexter and John T. Smith applied for paperhandler positions on October 20, 1994; they were not hired. Plaintiff Edwin Torres submitted an application for a paperhandler position to the Herald on February 13, 1995; he was not hired. At the time they applied, none of them knew anyone who worked at the Herald.

Thus, while the past history of nepotism is undisputed, how long the nepotistic practices continued, and what effect they may have had on these particular plaintiffs, is hotly contested. The question is not so much who the Herald hired and when, but how and why. Plaintiffs contend that the system of nepotism continues, and that the period from 1989 to the present should be understood as one continuing violation of Title VII. Defendant responds that the undisputed facts demonstrate that it has followed a policy neutral in intent and effect ever since early 1994. Whatever the theoretical faults of their prior system, they argue, it could not have adversely impacted plaintiffs who first applied to the Herald six months later.

The following charts reflect undisputed facts regarding the defendant's hiring of substitute paperhandlers from the time of the alleged change in hiring practices, some time after April 1994, until June 1996.

1. Hires between April 1994 and July 1996.

Hire Date Name Application Herald Application Race Date Relative Type ^fn_

2. September 2, 1994 Hiring

Plaintiff Gaines applied to the Herald on September 1, 1994. The Herald filled paperhandler position the next day, hiring Michael Child, a white man. There is little evidence of how it did so. Apart from its general assertions that it stopped hiring solely on the basis of nepotism sometime during the Grant case and after April 1994, the Herald is silent about the hiring procedure followed on September 2, 1994. The applicant who was hired, Michael Child, is one of the only successful applicants for whom no application is on file. Neither side has offered any evidence concerning whether he had friends and relatives at the Herald, when he applied, or which application form he used.

3. The Late September and October 1994 Hiring

The Herald next filled substitute paperhandler positions in late September 1994, hiring men who started work on September 29 and October 15, 1994. Although the evidence about the procedures followed at this time is much more extensive, several key questions remain unanswered. The Herald has submitted a list of applications that it believes were on file at that time, yet it acknowledges that the list is incomplete. Messing's affidavit states that the Herald had "at least 15 applications in its folder, ten of which were submitted earlier than David Gaines." It then offers the following list of applicants "of whom the Herald presently has a record": (emphasis added) Application Name "Insider"? Hired? Race/Ethni Date city 9/1/94 David Gaines No No black

Applications known to have been on file at Herald as of late September 1994

Nor does the record demonstrate conclusively that the date of the applications — however obtained — determined who was hired. One white applicant who had connections at the Herald was passed over, and two were hired. The record does not disclose why. The defendant acknowledges that its recordkeeping practices were shoddy; it cannot reconstruct all of the applications on file or how the calls were made. It is not even clear where the pool of applications was drawn from; for some of this time, unskilled applicants who did not specify "pressroom" and had no relatives in the pressroom would not have been forwarded there for consideration. See infra.

4. The July 15, 1995, Hiring

The next date on which defendant hired a substitute paperhandler was July 15, 1995. By this point, the Herald can account for all the applications but one, a Tim Buenopane, whose application was discarded because he did not live at the phone number he had listed. The applications on file at that time were as follows: Applications on File as of July 15, 1995 Application Date Name "Insider"? Race/Ethnicity 9/1/94 David Gaines No black 10/20/94 Frank F. Poindexter No black 10/20/94 John T. Smith No black 2/13/95 Edwin Torres No Spanish-surname d

The defendant has also submitted detailed evidence of the hiring procedure it followed in July 1995. It offers the contemporaneous handwritten list Daniel Messing made of his phone calls to applicants, recording the order in which he called them, and the result:

Applicant Called Result 7/95 Race/Ethnicity

John Smith (Name crossed "___working" white out) (crossed out) Tyrone Smith No Answer black Martin Vazquez No Answer Spanish-surnamed Anthony Ranzino No Answer white Glen Mercurio No Answer white Robert Kovalski No Answer white John Downey Wrong # white Tim Buenopane Don't [sic] Live unknown There Mark Miner No Answer black

David Gaines No Answer black Robert Baker No Answer unknown Bill Brothers No Answer unknown Steve Duplain No Answer unknown Jean Jean-Marie Wrong # black Dady Lochard No Answer black Keith Roberts No Answer black Francois Boston Disconnected black Frank Poindexter No Answer black John Smith T [sic] No Answer black Paul Sturtevant No Answer unknown

Messing's list stops there, before anyone has been reached or offered the position. In his affidavit, he adds that the first two applicants he reached and who were available were Michael McGrath and Richard Dondero, the twentieth and the twenty-first applicants he called. Both are white men, but both filled out one-sided, "outsider" applications. McGrath stated in his deposition that he did not know anyone at the Herald when he applied, but Dondero, the Herald concedes, did.

As evidence that the defendant did not, in fact, call the applicants in the order in which they applied, the plaintiffs offer the deposition testimony of Frank Poindexter and John T. Smith. Poindexter states that in July 1995, he had an answering machine which listed a beeper number, and that he neither received a message on his machine nor was he reached by beeper. John T. Smith also had voicemail in July 1995 and received no message from the Herald.

5. The June 1996 Hiring

Finally, the defendant offers evidence, in the form of Messing's affidavit, that another paperhandler position became open in June 1996. Again, Messing states that he called all of the applicants in the order in which they had applied, beginning with those still on file from the previous year. This time, he reached Tyrone Smith, Martin Vazquez, Anthony Ranzino, and Robert Kovalski. Martin Vazquez said he was unable to work at that time. The other three began work shortly thereafter. As with the 1994 hiring, Messing made no record of these calls. And again, one of the white males hired — Robert Kovalski — has on file a 2-sided application filled out in August 1994, long after the Herald contends it had stopped using such forms to screen initial applicants. The form lists a friend at the Herald.

6. Conclusion

Thus, there is a degree of uncertainty about how long nepotism continued to drive the pressroom hiring process. This uncertainty is heightened by Reilly's deposition testimony. In his deposition in July 1996, Reilly was less than clear about the change in the hiring procedures in the pressroom. "You usually go by the date of the applications that were made," he explained. Reilly Dep. at 23 (emphasis added). He asserted as well that this had been the practice since he became superintendent in 1989, id. at 31, 76, an assertion not only belied by the application records the Herald produced in the case, id. at 34-40, but also inconsistent with the Herald's claim that the pressroom policies underwent a sea change in early 1994. In addition, "going by the date of application" does not preclude nepotism, if only "insider" applications are kept on file. Finally, the plaintiffs offer the deposition testimony of Gaines, who states that when he called the Herald — on an unspecified date — to ask about the status of his application, the woman who answered the phone asked him whether he had any relatives who worked at the Herald. Deposition of David Gaines at 19.

D. Statistical Evidence

As additional support for their claims of disparate treatment and disparate impact, plaintiffs offer statistical evidence of an allegedly gross disparity between the racial composition of the Herald's work force and the percentage of qualified minority laborers in the local area. Because there are no specific qualifications for working as a paperhandler, the relevant labor market is that for able-bodied, unskilled workers. Plaintiffs offer the affidavit of Dr. Craig L. Moore, a professor of economics and statistics, who states that government data indicates that black or Spanish-surnamed minorities accounted for 15% of the Boston area labor market for "Handlers, Equipment Cleaners, Helpers Laborers" from 1988-1994. He concludes that the "actual proportion of minority employees in the pressroom is so small that the odds of it occurring by chance are virtually impossible." Although the defendant challenges several of Dr. Moore's statistical methods and assumptions, it is undisputed that from 1989-1995, the Herald employed only one or two black pressroom workers and no Spanish-surnamed pressroom workers. During the same period, the total number of pressroom employees ranged from 165 to 129.

E. The Herald's Knowledge of the Race of Pressroom Applicants

A final area of disputed fact is whether the Herald knew the race or ethnicity of pressroom applicants. Reilly testified in his deposition that he does not know anything about Boston neighborhoods that would enable him to deduce an applicant's race or ethnicity from his address. The plaintiffs respond that Reilly's testimony is hardly credible enough to preclude a factual dispute on the issue. For example, Reilly stated that he is not "bright" enough to recognize a Spanish surname: Edwin Torres is "probably" Spanish, and Mariel Rivera "possibly," but not Martin Vazquez or Juan Montero. Reilly Dep. at 74-75. In addition, two of the applicants — Jean-Paul Jean-Marie and Francois Boston — listed work or educational experience in Haiti on their applications. The race of "insider" applicants arguably could be deduced from the race of the friends and relatives they listed on their applications. Thus, the extent of the Herald's knowledge about the race of job applicants is still in dispute.

V. DISCUSSION OF THE DEFENDANT'S MOTION FOR SUMMARY JUDGMENT

The plaintiffs bring a number of counts under both federal and state law. Many of them raise similar claims or rely on identical evidence. I will address them accordingly.

A. Counts I III: Disparate Impact

Employment practices that are facially neutral and adopted without discriminatory intent may function to exclude minorities as effectively as intentional discrimination. Such practices may constitute a "disparate impact" violation of Title VII. 42 U.S.C. § 2000e-2(a)(2), 2000e-2(k)(1)(A)(a); Watson v. Fort Worth Bank Trust, 487 U.S. 977, 990, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988); Griggs, 401 U.S. at 432; E.E.O.C. v. Steamship Clerks Union, Local 1066, 48 F.3d 594, 601 (1st Cir. 1995), cert. denied, 516 U.S. 814, 116 S.Ct. 65, 133 L.Ed.2d 27 (1995). A policy or practice may be unlawful if it results in the selection of job applicants who differ significantly from the applicant pool. Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975).

There are three elements to a prima facie case of disparate impact: "identification, impact, and causation." Steamship Clerks, 48 F.3d at 601. The plaintiffs must identify either a particular employment practice of the defendant, or a group of practices that are not capable of separate analysis. 42 U.S.C. § 2000e-2(k)(1); Steamship Clerks, 48 F.3d at 601. They must then show that there is a disparate impact on members of the protected group to which they belong. Finally, they must link the practice and the disparate impact, showing that the former is the cause of the latter, Id.

Once the plaintiff has shown that a particular employment practice or group of practices has a disparate impact on the protected group to which they belong, the employer may avoid a finding of liability under Title VII by proving that the challenged practices are a "business necessity," in that they bear a significant relationship to successful job performance. 42 U.S.C. § 2000e-2(k)(1)(A)(i); Griggs, 401 U.S. at 436.

1. Disparate Impact

Because the law expects that nondiscriminatory hiring practices will, over time, produce a workforce whose composition mirrors that of the relevant community of qualified applicants, statistical evidence of "long-lasting and gross disparity between the composition of the workforce" and the relevant labor market may provide important proof of employment discrimination. Hazelwood School District v. United States, 433 U.S. 299, 307, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977), quoting International Bhd. of Teamsters v. United States, 431 U.S. 324, 340 n. 20, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977); Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 651, 109 S.Ct. 2115, 104 L.Ed.2d 733 (1989). Statistical evidence becomes more probative when combined, as it is here, with specific details of hiring practices and minority hiring rates during the relevant time period. See Steamship Clerks, 48 F.3d at 605; Latinos Unidos De Chelsea En Accion v. Secretary of Housing and Urban Dev., 799 F.2d 774, 785 (1st Cir. 1986).

There can be little dispute that the plaintiffs have shown a disparate impact on non-white applicants to the Herald. Substitute paperhandler positions require no particular skills or experience. The occupational category of "Handlers, Equipment Cleaners, Helpers Laborers" is clearly an appropriate point of comparison, although in a case involving unskilled labor, comparison to the general population would also have been probative. See Hazelwood, 433 U.S. at 308 n. 13. The plaintiffs offer statistical evidence that from 1988-1994, fifteen percent of Boston area workers in this category were black or Spanish-surnamed. The proportion of blacks and Hispanics in the general population may have been even higher. See Steamship Clerks, 48 F.3d at 604 (reporting that blacks and Hispanics made up 27% of the available workforce in Boston in 1991).

By either measure, the Herald's workforce was disproportionately white. Yet out of the 134 pressmen who worked at the Herald in 1994, only one was black and none were Spanish-surnamed. Nor was this picture being gradually altered by more balanced hiring decisions made during this time. Of the nine paperhandlers hired in 1994 and 1995, all were white. Only in June 1996, does the racial composition of the pressroom begin to shift: of three paperhandlers hired, one was black, while a Spanish-surnamed applicant was offered and turned down a position. Thus, the plaintiffs have shown a disparate impact on black and Spanish-surnamed applicants at least until mid-1996.

2. Identified Employment Practices that Fail to Make Out a Disparate Impact Claim

a. Practices That Could Not Have Affected these Plaintiffs

Having demonstrated a disparate impact, the plaintiffs must link it to a particular employment practice or group of practices. Plaintiffs have identified a laundry list of practices that they allege have a discriminatory impact on black and Spanish-surnamed applicants to the Herald pressroom. They have also provided statistics that show that the Herald's workforce has long been almost overwhelmingly white. A number of these identified practices, however, fail to support a prima facie disparate impact claim because they did not cause a disparate impact on the plaintiffs themselves.

Several practices the plaintiffs identify could not have affected anyone not already at the Herald or with contacts there, and thus cannot have affected the plaintiffs personally. Plaintiffs do not have standing, either individually or as representatives of a class, to challenge policies that did not affect them. See Warth v. Seldin, 422 U.S. 490, 501, 95 S.Ct. 2197, 45 L.Ed.2d 343 (1975); East Texas Motor Freight System, Inc. v. Rodriguez, 431 U.S. 395, 403-04, 97 S.Ct. 1891, 52 L.Ed.2d 453 (1977); Donnelly v. Rhode Island Bd. of Governors for Higher Educ., 110 F.3d 2, 4 (1997). They cannot challenge the failure to post job openings on Herald premises, where non-employees would not have seen them, or to maintain written promotion policies, which could only have affected existing employees.

The lack of any possible impact on the plaintiffs also precludes their challenge to the system of promoting full-time paperhandlers exclusively from within the pool of current substitutes. This specific practice definitely would have prevented the plaintiffs from obtaining the full-time positions for which some of them applied, had there been any such positions available. However, between the time they first applied to the Herald and the time of filing this lawsuit, there were no full-time paperhandler positions available: no full-time paper handler positions were filled between June 5, 1994, three months before Gaines applied, and July, 1996.

b. The Defendant's Failure to Take Steps to Create a More Diverse Workforce

Plaintiffs also identify as discriminatory the defendant's failure to take affirmative steps to recruit and hire black and Spanish-surnamed applicants. The Herald, they point out, does not: advertise pressroom vacancies to the public; "provide an effective application process" for minority applicants; inform minority applicants of vacancies; or solicit or recruit minority applicants. Title VII, however, does not require any employer to take such affirmative steps until after a violation has been found. 42 U.S.C. § 2000e-2(j); Furnco Constr. Corp. v. Waters, 438 U.S. 567, 578, 98 S.Ct. 2943, 57 L.Ed.2d 957 (1978).

3. The Nepotism Claim

The plaintiffs are more successful in their challenge to the Herald's practice of giving preference in hiring to friends and relatives of current employees, who filled out two-sided applications they obtained from Reilly directly. As the First Circuit noted in Grant, when a workforce is predominantly white, nepotism can operate to exclude applicants on the basis of race. Grant, 55 F.3d at 8; accord Steamship Clerks, 48 F.3d at 606; Thomas v. Washington County Sch. Bd., 915 F.2d 922, 925 (4th Cir. 1990); Bonilla v. Oakland Scavenger Co., 697 F.2d 1297, 1303 (9th Cir. 1982), cert. denied, 467 U.S. 1251, 104 S.Ct. 3533, 82 L.Ed.2d 838 (1984); United States v. International Ass'n of Bridge, Structural and Ornamental Iron Workers, Local No. 1, 438 F.2d 679, 683 (7th Cir. 1971), cert. denied, 404 U.S. 830, 92 S.Ct. 75, 30 L.Ed.2d 60 (1971); Garland v. USAir, Inc., 767 F. Supp. 715, 723 (W.D.Pa. 1991).

The employer's own records, demonstrating that all of those hired were relatives of incumbent employees, may be sufficient to prove that the policy of nepotism caused the disparate impact. See Steamship Clerks, 48 F.3d at 606. The defendant's argument that passive reliance on word-of-mouth recruiting is not actionable under Title VII, see EEOC v. Chicago Miniature Lamp Works, 947 F.2d 292, 305 (7th Cir. 1991), is beside the point, given the evidence that the pressroom hiring practices — from the distribution of two different application forms, to the information requested on those forms, to the selection process itself — actively favored applicants with connections at the paper.

The defendant does not counter the evidence of nepotism with any argument that these practices were a business necessity, obviating the need for any analysis on that score. They argue simply that they now employ an entirely neutral system, and have done so since before any of the plaintiffs applied.

As detailed above, whether and when this nepotistic system ended is still seriously in dispute. There is evidence in the Grant deposition testimony, reaffirmed in depositions in this case, that it continued into 1994. The application forms submitted in this case reveal that as late as September 1994, job applicants with friends and relatives at the Herald were still receiving and submitting two-page "insider" applications. Two of the three paperhandlers hired in September and October 1994 were among them. There is no application on file for the third. Gaines reported that when he inquired about his application he was asked whether he had relatives at the Herald.

To defeat this evidence of nepotism, defendant offers generalized assertions by Messing and Reilly that they ceased using a nepotistic system some time during 1994, after the Otis Grant suit, and began calling applicants according to the order of their application. Such vague assertions are not sufficient to support, or defeat, a motion for summary judgment. Teamsters, 431 U.S. at 342 n. 24. The only detailed and specific evidence the defendant offers is Messing's affidavit. That affidavit is silent on how the paperhandler who began work on September 2, 1994, was hired. For the hiring decisions made at the end of September, defendant offers Messing's account that he called all of the applicants in the order in which they applied. However, he does not purport to have either a record or a recollection of all of the applications on file or the order in which they were called. In addition, even the list of applications he offers reflects that at least four of the paperhandlers with applications earlier than that of David Gaines had submitted "insider" applications. Two men whose applications predated Gaines' were not hired, casting doubt, to a degree, on the "order of application" claim. Thus, there is enough factual dispute about the continuation and effect of the Herald's nepotistic practices through October 1994 for the plaintiffs to survive the defendant's motion for summary judgment.

The evidence the Herald offers concerning the July 1995 hiring is somewhat clearer. In addition to Messing's affidavit, it offers his contemporaneous list of which applicants he called and in what order. The list demonstrates that the applicants were in fact called in the order in which they had applied. The two people reached and hired, although white, had filled out one-page applications. They also came lower on the list than three of the four plaintiffs, eliminating any suggestion that they had been placed higher on the list as a result of nepotism.

Plaintiffs evidence that the Herald had not abandoned the nepotistic system is slim: Dondero's Herald connections, the fact that Messing's list leaves off before he reached either of the men hired, and two plaintiffs' claims that they had voicemail but received no message from the Herald. Drawing every favorable inference in favor of the plaintiffs, there is no evidence that Messing had any knowledge that Dondero knew anyone at the Herald until after Dondero was hired. As to the failure to leave messages on two of the plaintiffs' machines, Messing never claimed that the first-reached-first-hired procedure he followed involved leaving messages for applicants who were not at home. The objective record shows that the Herald needed one paperhandler to start immediately: Michael McGrath started that same day. Why the list leaves off where it does, one can only speculate.

The June 1996 hiring came after the date that this lawsuit was filed. It is worth addressing nonetheless, because the plaintiffs raise a claim that the Herald's past and present hiring practices constitute a single continuing violation of Title VII. The defendant offers detailed evidence that in this case Messing again filled a paperhandler opening by calling all applicants in the order in which they had applied. Messing's affidavit that he did so is corroborated by the fact that the four men offered jobs were among the earliest five applicants on file. One of them was black, one Spanish-surnamed, and two were white. The fact that the fifth name on the list was that of Robert Kovalski, who had filled out a two-page application listing contacts at the Herald, does not raise a material issue about whether this decision was made in accordance with order of application.

Thus, there are material facts in dispute with regard to whether the Herald continued at least through October 1994 to use hiring practices that operated to exclude applicants from employment because of their race. It is disputed whether the nepotistic system was still in place at that time, especially with regard to the September 2 hire, about which the defendant has offered no evidence, but also with regard to the September and October decisions. With regard to the positions filled in July 1995 and June 1996, however, the defendant has put forward detailed evidence demonstrating that it no longer relied on nepotism. The plaintiffs have been unable to counter this evidence.

4. Employment Practices Related to the Nepotism Scheme

Plaintiffs also identify a number of practices that together made the nepotistic scheme possible: the all-white composition of the pressroom; the pressroom superintendent's unchecked, unfettered discretion in making hiring decisions; the failure to contact applicants in writing or to maintain formal criteria for paperhandler hiring decisions, either in writing or otherwise; and the failure to keep records of job applicants or their race. Although many of these practices, standing alone, might not make out a Title VII violation, together they form the background to the informal nepotistic system at the heart of this case.

An "undisciplined system of subjective decisionmaking" can have the same effects on minority applicants as an intentionally discriminatory system, and it can constitute a violation of Title VII under either a disparate treatment or a disparate impact theory. Watson v. Fort Worth Bank Trust, 487 U.S. 977, 990-91, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988). Such a violation can be made out when, as here, there is evidence that the subjective decision-making process coincides with a "significant disparity in the representation of a particular group." Mohammed v. Callaway, 698 F.2d 395, 401 (10th Cir. 1983).

The plaintiffs have introduced evidence that hiring paperhandlers is "entirely up to" the pressroom superintendent, without guidance from any formal criteria, written standards, or job offer procedures. Reilly Dep. at 21-24. As mentioned above, the statistical evidence shows that this subjectivity was combined with a gross racial disparity between the pressroom workforce and the local community. Again, however, the defendant offers evidence that the pressroom superintendent's decisions are no longer subjective, but instead are governed by a simple, neutral system based on application date. The parties' motions with regard to these challenged practices — whether under the disparate impact or disparate treatment theory — therefore must follow my analysis of the nepotism claim.

5. Conclusion as to the Disparate Impact Claims

Defendant's motion for summary judgment on Counts I and III is DENIED with respect to the September and October 1994 hiring decisions and with respect to Gaines. It is ALLOWED with respect to the July, 1995 hiring decision and with regard to plaintiffs Poindexter, Smith, and Torres.

B. Count IV: Inquiry About Race in the Application Process, in Violation of Massachusetts Law

The plaintiffs also attack the use of the two-sided form on state law grounds, arguing that it "expresses, directly or indirectly . . . limitation, specification, or discrimination" as to race, or an intent to discriminate on the basis or race. Mass. Gen. L. ch. 151B, § 4(3). They also allege that the Herald impermissibly inquired about race in connection with employment. Id.

The caselaw interpreting this statute is scarce, see Morgan v. Hennigan, 379 F. Supp. 410, 456-57 (Mass. 1974) (referring in passing to the city of Boston's belief that the law requires applications that do not refer to race), and the parties have provided little guidance. Plaintiffs merely state that the facts show there is a genuine issue to be tried, while the defendant asserts that its arguments with regard to the Title VII claims dispose of this as well.

The language of the statute prohibits any inquiry with regard to employment that indicates the employer's intent to specify a preferred or required race, classify applicants according to race, or, more generally, discriminate on the basis of race. This intent can be expressed directly or indirectly. The Massachusetts Commission Against Discrimination ("MCAD"), pursuant to its statutory authority to implement and enforce the state's anti-discrimination laws, Mass. Gen. L. ch. 151B, § 3(5), has issued guidelines prohibiting any "inquiry into the race or color of an applicant." Mass. Regs. Code tit. 804, § 3.02. Nor may an employer make less direct inquiries "the response to which would likely disclose the applicant's protected status." Mass. Regs. Code tit. 804, § 3.01.

Violations of this law can give rise to claims by those directly asked prohibited questions. See Kraft v. Police Comm'r of Boston, 410 Mass. 155, 571 N.E.2d 380, 382 (1991) (involving the analogous Mass. Gen. L. ch. 151B, § 4(9A), which prohibits inquiry about mental health). Under the analogous federal statute, 42 U.S.C. § 2000e-3(b), suits have been brought by those with a genuine interest in the job who are "effectively deterred" from applying, see Hailes v. United Air Lines, 464 F.2d 1006, 1008 (5th Cir. 1972), or who have been confronted with discriminatory advertising in the course of their search for employment. See Morrow v. Mississippi Publishers Corp., 1972 WL 236, *2 (S.D.Miss.). Impermissible advertisements have also been taken as evidence of discriminatory motivation. See Capaci v. Katz Besthoff, Inc., 711 F.2d 647, 659 (5th Cir. 1983), cert. denied, 466 U.S. 927, 104 S.Ct. 1709, 80 L.Ed.2d 182 (1984).

Given the overall composition of the Herald's workforce, a request for names of friends and relatives employed at the Herald could constitute an inquiry the response to which would likely disclose an applicant's protected class status. The more difficult question is whether the plaintiffs were actually subjected to an inquiry about their race. Only Gaines alleges that the Herald inquired about his relatives, in the phone call he made to check on the progress of his application. None of the plaintiffs were given the application form that asks for names of friends and relatives at the paper.

The claim in this case is a rather unusual one: that although the plaintiffs were not asked about their race, others who were asked benefitted from responding in a way that showed they were white. Viewing this claim more abstractly, it is a claim that the Herald inquired about race in the application process, and the plaintiffs were injured as a result. As persons injured by an act that violated state anti-discrimination laws, the plaintiffs should be able to seek redress.

The facts supporting this claim mirror those supporting the disparate impact nepotism claim. The application forms submitted in this case show that the form requesting information on friends or relatives at the Herald continued to be submitted at the initial application stage through the Fall of 1994. Robert Kovalski submitted such an application on August 25, 1994, and Robert Baker did so on September 8, 1994. From then on, the initial applications on file in the pressroom were all one page applications, with the exception of Brian Kelly's application, submitted on April 6, 1995. As discussed above, however, the employment decisions made in July 1995 were based on order of application, with no evidence of preference to applicants who had listed their friends and relatives at the Herald. Thus, the only persons who could have been injured by this inquiry are those who applied before October 15, 1994, the last date on which an employment decision could have been made on the basis of application forms that inquired about race.

The defendant's motion for summary judgment on Count IV is therefore DENIED with respect to Gaines and with respect to hiring decisions before October 15, 1994. It is ALLOWED with respect to subsequent hiring decisions and with respect to plaintiffs Smith, Poindexter, and Torres.

C. Count V, Count VI, Count VII: Intentional Discrimination

Plaintiffs bring four claims of intentional discrimination, alleging that the defendant: discriminated against them by failing or refusing to hire them because of their race in violation of 42 U.S.C. § 2000e-2(a)(1); denied them the same right to make and enforce contracts as is enjoyed by white citizens, 42 U.S.C. § 1981; and refused to hire them on the basis of their race, in violation of Massachusetts law. Mass. Gen. L. ch. 151B, § 4(3). A single framework applies to the analysis of all three claims. See Oliver v. Digital Equip. Corp., 846 F.2d 103, 111 (1st Cir. 1988) (Title VII analysis of intentional discrimination claims applies to Section 1981); T S Servs. Assocs., Inc. v. Crenson, 666 F.2d 722, 724 (1st Cir. 1981) (same), and Mass. Gen. L. ch. 151B, Serafino v. Hasbro, 82 F.3d 515, 518-19 (1st Cir. 1996) (same analysis applies to Mass. Gen. L. ch. 151B); Matthews v. Ocean Spray Cranberries, Inc., 426 Mass. 122, 686 N.E.2d 1303, 1308 (1997) (same); Wheelock College v. Massachusetts Comm'n Against Discrimination, 371 Mass. 130, 355 N.E.2d 309, 314-15 (1976).

The burden of making out a prima facie case of disparate treatment is "not onerous." Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 253, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981). A plaintiff establishes a prima facie case of disparate treatment in hiring by demonstrating that: he is a member of a protected class; that he applied and was qualified for a position for which the employer was seeking applicants; he was not considered for or was rejected for this position; and the position was kept open and the employer continued to seek applicants with the same qualifications. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973).

Once the plaintiff makes out a prima facie case, a presumption arises that the employer unlawfully discriminated against him. The defendant must then rebut the presumption by producing admissible evidence of a legitimate, nondiscriminatory reason for the hiring decision in question. St. Mary's Honor Center v. Hicks, 509 U.S. 502, 510-11, 113 S.Ct. 2742, 125 L.Ed.2d 407 (1993); Burdine, 450 U.S. at 255. If it does so, the burden shifts back to the plaintiff to show that this reason is a pretext, masking discriminatory intent. Hicks, 509 U.S. at 515.

To survive summary judgment at this stage, the plaintiff must come forward with evidence sufficient to permit a rational trier of fact to conclude two things: first, that the reason offered is pretextual, and second, that it is a pretext for discrimination. Udo, at 13. The plaintiff may rely on the same evidence to prove both pretext and discriminatory animus. Smith, 40 F.3d at 16.

The undisputed facts in this case establish the basic elements of a prima facie case of disparate treatment: plaintiffs are members of a protected class and were qualified for the paperhandler positions at issue, positions that did not require any specific work or educational experience. They were not hired, and the positions were given to white applicants with the same qualifications.

The defendant comes forward with two arguments in rebuttal: first, that intentional discrimination would have been impossible, because the Herald had no way of knowing the race or ethnicity of any pressroom applicants; second, that there was a legitimate, non-discriminatory reason for its failure to offer plaintiffs the jobs: calling the applicants in the order in which they had applied, Messing reached and offered the job to someone else first.

Both arguments are inextricably bound up with the issue of whether and when the Herald ceased relying on nepotism. Knowledge of an applicants' race could have been derived from knowledge of the race of his friends and relatives already working at the Herald. When the defendant stopped relying on the two-page applications is thus as central to this issue as to the issue of disparate treatment. Among the plaintiffs' one-page applications, however, only Edwin Torres' gave any indication as to ethnic background. Even Reilly admitted that Torres was "probably" a Spanish surname. Because Torres did not apply to the Herald until February, 1995, the materiality of this fact turns on the strength of the defendant's proof that it had by then turned to a purely neutral hiring system.

As noted above, the defendant has failed to provide any evidence regarding the September 2, 1994, hire, and the evidence it offers regarding the September 29 and October 15 hires is inconclusive. In both cases, a rational trier of fact could conclude that the Herald had not yet abandoned its nepotistic hiring system, a system that — given the all-white composition of the workforce, and Reilly's somewhat hard-to-credit claims that he cannot recognize a Spanish surname, a black neighborhood, or the probable race of his current employee's relatives — a reasonable fact-finder could conclude masked discriminatory animus. With regard to the July 1995 hire, however, the plaintiffs have produced little evidence to raise a triable issue over whether the defendant failed to call the applicants in the order in which they applied, or that to do so was in fact a pretext for discrimination.

Thus, as with the disparate impact claims, defendant's motion for summary judgment on the intentional discrimination claims, Counts V, VI, and VII, is DENIED with respect to the September and October 1994 hiring decisions and with respect to Gaines. It is ALLOWED with respect to the July 1995 hiring decision and with regard to plaintiffs Poindexter, Smith, and Torres.

D. Count II: Mixed Motive Discrimination

Plaintiffs also bring a claim based on 42 U.S.C. § 2000e-2(m), which makes unlawful an employment practice in which race "was a motivating factor, even though other factors also motivated the practice." Although the plaintiffs' complaint refers both to the Herald policy of "reserving all full-time paperhandler positions exclusively for white males" and, more generally, to "unlawful selection and hiring of pressroom employees," their argument on this Count reveals that it is directed at the policy of hiring full-timers exclusively from among substitute paperhandlers. As discussed above, this Count must fail because there were no full-time paperhandler positions available between the time the first plaintiff applied to the Herald and the time this suit was filed. The defendant's motion for summary judgment on Count II is ALLOWED.

E. Count VIII: Violation of Mass. Gen. L. ch. 93, § 102(2)

Plaintiffs' final count arises under Mass. Gen. L. ch. 93, § 102, which guarantees that all citizens of Massachusetts "shall have . . . the same rights enjoyed by white male citizens, to make and enforce contracts. . . ." Chapter 151B, however, is the exclusive state law remedy for employment discrimination claims. Woods v. Friction Materials, Inc., 30 F.3d 255, 264 (1st Cir. 1994); Charland v. Muzi Motors, Inc., 417 Mass. 580, 631 N.E.2d 555, 559 (1994). The defendant's motion for summary judgment on Count VIII is ALLOWED.

VI. THE PLAINTIFFS' MOTIONS FOR SUMMARY JUDGMENT

Plaintiffs have moved for summary judgment on Counts I, II, III, and VIII. My grant of summary judgment to the defendants on Counts II and VIII (mixed motive discrimination in the hiring of full-time paperhandlers and violation of Mass. Gen. L. ch. 93, § 102(2)) precludes granting summary judgment to the plaintiffs on those counts. As the preceding discussion outlines, there are still important factual disputes surrounding Counts I and III, which challenge the Herald's nepotistic system. The defendant has offered evidence in the form of deposition testimony and Messing's affidavit that the use of the two-sided forms and the practice of preferring applicants with friends and relatives at the Herald ended some time in the spring of 1994. It is undisputed that Gaines applied to the Herald on September 1, 1994. Drawing every inference in favor of the defendants, it is possible that a reasonable factfinder could find that the Herald had adopted a neutral hiring system by the fall of 1994 and that the three men who were hired at that time were simply the earliest applicants who could be reached by phone. The plaintiffs' motion for partial summary judgment on Counts I, II, III, and VIII is DENIED.

VII. THE PLAINTIFFS' MOTION FOR CLASS CERTIFICATION

The plaintiffs have sought to certify a broad class: all black and Spanish-surnamed past, present, and future applicants to the Herald; those who were deterred from applying; and those who were not informed of openings in the pressroom or were otherwise not recruited by the Herald. My grant of summary judgment to the Herald on the failure-to-recruit claim and with regard to hiring decisions after October 15, 1994, precludes claims brought on behalf of present and future applicants and those who were not recruited. The remaining potential class members are past black and Spanish-surnamed applicants to the Herald and those who were deterred from applying.

The plaintiffs have offered significant evidence that the defendant acted in a way generally applicable to the proposed class of black and Spanish-surnamed applicants, such that any injunctive relief awarded would appropriately cover the class as a whole. Fed.R.Civ.P. 23(b)(2). In order for this case to proceed as a class action, however, all of the requirements of Rule 23 must be met. General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). Rule 23 sets out four requirements for a class action:

(1) the class is so numerous that joinder of all members is impracticable;

(2) there are questions of law or fact common to the class;

(3) the claims of the representatives are typical of the claims of the class; and

(4) the representative parties will fairly and adequately represent the interests of the class. Fed.R.Civ.P. 23.

There is no dispute that there would be common questions of law or fact in the proposed class, nor that the plaintiffs could fairly and adequately represent the other class members. Thus, class certification in this case turns primarily on the practicability of joinder, with the scope of the class depending on the typicality of the claims of the one remaining named plaintiff, David Gaines.

A. The Typicality of Gaines' Claims

Gaines applied to the Herald and was not hired — either explicitly because of his race, or because of a policy of nepotism that acted to exclude non-white applicants. His claims are thus typical of those who applied to the Herald and were not hired, either intentionally or indirectly because of their race. His claims are not typical of the claims of proposed class members who were discouraged from applying altogether. Thus, he can only represent the claims of the first group. See Falcon, 457 U.S. at 158.

The plaintiffs have moved to certify a class that includes both black and Spanish-surnamed applicants, but I have granted summary judgment to the defendant on the claims of the one Spanish-surnamed plaintiff, Edwin Torres. Nonetheless, Gaines' claims are typical of those of Spanish-surnamed as well as black applicants. The evidence in this case suggests that all non-white applicants without friends and relatives at the Herald were excluded by the same hiring practices and decisions. The defendant's nepotistic hiring system made no meaningful distinction, either in its motivation or its effect, between members of different minority groups. See Carpenter v. Stephen F. Austin State Univ., 706 F.2d 608, 617 (5th Cir. 1983). Should evidence arise of significantly different treatment of these two groups and of possibly conflicting interests, the proposed class could later be subdivided. See Majeske v. City of Chicago, 740 F. Supp. 1350, 1357 (N.D.Ill. 1990).

B. Temporal Boundaries of the Class Claims

The first step in determining the practicability of joinder is to determine how many potential class members there are. Excluding those who never applied, the plaintiffs' proposed class would include all past, present, and future black and Spanish-surnamed applicants to the Herald pressroom who were not hired. My grant of summary judgment to the defendant with regard to hiring decisions made from July 1995 onwards disposes of potential claimants who applied any time after the last arguably nepotistic hiring decision, that is, any time after October 15, 1994, when John A. Smith began work.

The size of the class thus turns on how far back in time the proposed class can extend, which, in turn, is a statute of limitations issue. Although not every member of a class of plaintiffs must have filed a claim with the EEOC or MCAD, the class is limited to those whose causes of action were alive on the date the first named plaintiff filed. Albemarle, 422 U.S. at 414 n. 8; Beasley v. Griffin, 81 F.R.D. 114, 118 (Mass. 1979). This is a question first of the applicable statutes of limitations for the plaintiffs' several claims, and second, on any legal theories that would extend or toll those limitations.

There are three separate statutes of limitations at issue here: for Title VII, for the state law claims, and for Section 1981. In Massachusetts, a Title VII claim is timely if filed with the Massachusetts Commission Against Discrimination ("MCAD") within 240 days of the complained of act of discrimination. 42 U.S.C. § 2000e-5(e)(1); Sabree v. United Bhd. of Carpenters and Joiners, Local No. 33, 921 F.2d 396, 399 (1st Cir. 1990). Claims under M.G.L. ch. 151B must be filed within six months. M.G.L. ch. 151b § 5; Lawton v. State Mut. Life Assurance Co. of Am., 924 F. Supp. 331, 339 (Mass. 1996), aff'd 101 F.3d 218 (1st Cir. 1996). Section 1981 claims borrow the state law statute of limitations for personal injury actions, which in Massachusetts is three years. Johnson v. Rodriguez, 943 F.2d 104, 107 (1st Cir. 1991), cert. denied, 502 U.S. 1063, 112 S.Ct. 948, 117 L.Ed.2d 117 (1992).

1. The Title VII Claims

Gaines was the first named plaintiff to file an administrative complaint in this case. He filed his complaint with the MCAD on February 17, 1995. Under normal circumstances, he could not bring a Title VII claim based on any acts of discrimination that occurred more than 240 days before, that is, before June 22, 1994. Between that time and the defendant's switch to the current neutral order-of-application hiring policy, only three pressroom positions were filled: on September 2, September 29, and October 15, 1994.

What, then, is the size of this potential class of black and Spanish-surnamed applicants who had applied but were not considered for the September and October 1994 paperhandler positions because they had no friends or relatives at the Herald? The Herald maintains that it has kept all applications it received since some time in early 1994. As discussed above, however, that contention is still a matter of dispute; not only are all the applications on file from early 1994 two-sided "insider" applications, but one black applicant to the Herald stated at his deposition that he had applied in 1994 — but there is no record of his application. Deposition of Terrell Johnson. Theoretically, any application submitted since 1987 — the date the plaintiffs propose for the beginning of their class — could have still been on file in September 1994.

The defendant at one point alleged that it had a policy of discarding all unsuccessful applications after twelve to eighteen months, but this allegation is contradicted by the Herald's own employment records: Daniel Savage, hired on September 29, 1994, applied on March 3, 1992; Dennis Messing, hired on September 8, 1990, applied on November 12, 1988, and Gary Miller, hired on March 29, 1992, applied on October 17, 1989. The outside date for hiring from an old application seems, then, not to be eighteen months but two to three years.

There is an additional uncertainty about whether "outsider" applications were kept at all. In his deposition in the Grant case, on April 7, 1994, Reilly stated that the only application form that had been used to hire substitute paperhandlers was the two-sided form. It is not clear whether this means that "outsider" applications were immediately discarded, or that they were kept on file and never called. If they were discarded immediately, the act of discrimination was this initial rejection of their application, not a later hiring decision made from a list from which they had long been removed. See United Air Lines v. Evans, 431 U.S. 553, 558, 97 S.Ct. 1885, 52 L.Ed.2d 571 (1977). That would put the claim of anyone who applied prior to June 22, 1994, outside the limitations period. If they were kept on file without being called, however, each hiring decision for which their applications were on file but were ignored would constitute another act of discrimination. This would mean that potentially anyone who had applied in the three years prior to September 2, 1994, is a potential class member.

The plaintiffs seek to extend the proposed class further back in time, on the theory that the nepotistic hiring system constituted a "continuing violation" of Title VII, and that therefore the statute of limitations is inapplicable. See, e.g., Kassaye v. Bryant College, 999 F.2d 603, 606 (1st Cir. 1993). This Circuit recognizes two forms of continuing violations, serial and systemic. Jensen v. Frank, 912 F.2d 517, 522 (1st Cir. 1990). A serial violation is made up of a number of discriminatory acts, each arising from the same animus, such as a single employee being repeatedly passed over for a promotion. Id.; Mack v. Great Atlantic and Pacific Tea Co., 871 F.2d 179, 183 (1st Cir. 1989). A systemic violation is a continuing illegal policy, rather than a specific act. Mack, 871 F.2d at 183.

The continuing violation doctrine allows plaintiffs to bring suits on the basis of later acts committed within the limitations period, even though claims based on the earlier acts or manifestations of the policy had "grown stale." Mack, 871 F.2d at 183; Velazquez v. Chardon, 736 F.2d 831, 833 (1st Cir. 1984) ("if the statutory violation occurs as a result of a continuing policy, itself illegal, then the statute does not foreclose an action aimed at the company's enforcement of the policy within the limitations period") (emphasis added). Thus, a plaintiff is not barred from bringing a claim based on discriminatory acts or policies that extended into the limitations period, simply because there had been earlier, now-time barred, manifestations of the same animus or policy. The continuing violation theory thus does not help the plaintiffs to expand their claim to cover the Herald's policies or hiring decisions prior to June 1994.

The potential class, then, would include at its most expansive only black and Spanishsurnamed applicants for unskilled positions at the Herald who had applied after September 2, 1991. Six applications of potential class members are in the defendant's files: those of Tyrone Smith, Martin Vasquez, Jean-Paul Jean-Marie, Mark Miner, plaintiff David Gaines, and Dady Lochard. The number of earlier applications beyond these six can only be deduced from the facts before me as to more recent applications. See Westcott v. Califano, 460 F. Supp. 737, 744 (Mass. 1978), aff'd 443 U.S. 76, 99 S.Ct. 2655, 61 L.Ed.2d 382 (1979).

The plaintiff has submitted a list of seventy black and Spanish-surnamed applicants who applied for an unskilled position at the paper between July 1994 and May 1996. Three are duplicates, leaving a total of sixty-seven. This number may be unusually high, due to rumors of this very lawsuit, rumors that prompted the applications of two of the plaintiffs. Counterbalancing this possible source of over-counting, however, is the fact that the race of many applicants has not yet been identified; of the approximately 170 surveys the plaintiffs sent out to Herald applicants, only sixty were returned with answers indicating that the applicant was either white, black, or "Hispanic Surnamed."

In addition, there is considerable confusion about how many of these applications are typical of those that would have been forwarded to the pressroom during the relevant time period. The defendant's Vice President of Human Resources, David St. Mary ("St. Mary"), states in an affidavit that as a general rule, the Herald does not forward all unskilled applicants to the pressroom for consideration as paperhandlers. Eliminating the thirty-two applications that do not specifically request paperhandler or pressroom positions would yield a total of thirty-five; eliminating the applicant who requested only clerical positions would leave thirty-four.

In his deposition, however, Reilly stated that applications for unspecified unskilled positions were in fact forwarded to the pressroom. Reilly Dep. at 95. Evidence from the applications themselves is inconsistent: while some general applications do not appear in pressroom records, others do. To confuse matters further, three black applicants who listed "paperhandler" or "pressman" on their applications are inexplicably included in St. Mary's list of applications that were not forwarded to the pressroom because they failed to "indicate an interest in the pressroom." (Roy McDonald, Oral Miller, and Christopher Jones).

Extrapolating potential class members from these conflicting figures is an uncertain business. At this point, I will err on the side of inclusiveness, well aware that if further proceedings in this case resolve some of the factual ambiguities over whether outside applications were thrown out immediately, or how often unskilled applications were forwarded to the pressroom, this number might be significantly reduced. It might then be necessary to reexamine the issue of class certification.

Accepting for the time being the plaintiffs' list of black and Spanish-surnamed applicants, it shows ten applicants for the last five months of 1994, thirty-eight applicants in 1995, and seventeen applicants for the first five months of 1996, for an average of twenty-five to forty a year. This would yield a potential class of seventy-five to 120 applicants for the three years from September 2, 1991, to October 15, 1994, if the applications were kept but ignored. If they were immediately rejected, the class would only be those who had applied between June 22, 1994, and October 15, 1994.

2. The 151B, § 4 Claims

The same analysis applies to the 151B, § 4 claims, with the one difference that the limitations period is six months, rather than 240 days. The potential plaintiffs would thus be those who had applied to the Herald between August 17, 1994, and October 15, 1994, if all outsider applications were discarded as soon as they were received. If they were kept but disregarded, the potential class would again be all of those who had applied between September 2, 1991, and October 15, 1994.

3. The Section 1981 Claim

The three-year statute of limitations under Section 1981 simplifies the identification of the potential claimants under this theory. The complaint in this case was filed on August 29, 1995. Any black or Spanishsurnamed applicant to the pressroom who applied and was rejected between August 29, 1992, and October 15, 1994, would be included, whether their applications were immediately rejected or kept on file but ignored. The Herald's records show that there were sixteen substitute paperhandler positions filled during this time. The pool of applicants, using the previous figure of twenty-five to forty a year, would be again seventy-five to 120.

C. The Certified Class

Although numbers alone do not dictate if the class is sufficiently numerous that joinder is impracticable, Andrews v. Bechtel Power Corp., 780 F.2d 124, 131 (1st Cir. 1985), cert. denied 476 U.S. 1172, 106 S.Ct. 2896, 90 L.Ed.2d 983 (1986), joinder of as many as seventy-five to 120 potential plaintiffs would be impracticable at this stage. See, e.g., Beasley, 81 F.R.D. at 116 (certifying class where 64 employees had been identified, ten of whom had opted out of the class). Other factors that may affect whether joinder is impracticable include whether all class members live in the same geographic area, and whether they can be easily identified. Id. at 132. Although presumably many of the class members do live in the surrounding area, the absence of complete application records makes them hard to identify. See Padilla v. Stringer, 395 F. Supp. 495, 503 (N.M. 1974) (certifying a class in part because records of unsuccessful applicants had been destroyed). Under the circumstances, I will certify an appropriate class, mindful of the fact that subsequent evidence introduced during the course of this litigation may require a reconsideration of its dimensions. See Falcon, 457 U.S. at 160.

The class thus certified consists of:

1. For the Title VII and Mass. Gen. L. ch. 151B, § 4 claims:

All black and Spanish-surnamed applicants for unskilled positions at the Herald who applied between September 2, 1991, and October 15, 1994, and were not hired.

2. For the Section 1981 claim:

All black and Spanish-surnamed applicants for unskilled positions at the Herald who applied between August 29, 1992, and October 15, 1994.

VIII. CONCLUSION

For the foregoing reasons, and upon review of the pleadings and the memoranda and affidavits submitted therewith, the plaintiffs motion for partial summary judgment (docket # 24) is DENIED. The defendant's motion for summary judgment (docket # 19) is GRANTED on Counts I, III, IV, V, VI, and VII with respect to the claims of Frank Poindexter, John T. Smith, and Edwin Torres, and with respect to employment practices after October 15, 1994, but DENIED with respect to the claims of David Gaines and with respect to employment practices prior to that date. The defendant's motion for summary judgment on Counts II and VIII is ALLOWED. The plaintiffs' motion for class certification (docket # 6) is ALLOWED, as modified by this memorandum and the accompanying order. The plaintiffs' assented to motion to file the defendant's EEO-1 reports for the years 1992 through 1995 (docket # 48) is GRANTED. Plaintiffs' motion for leave to file second affidavit of expert witness Craig L. Moore (docket # 49) is MOOT. Defendant's motion to strike the affidavit of Anthony Neal (docket # 35) is DENIED. Plaintiffs' motion to remove Samuel Cheatem as a named plaintiff (docket # 34) is GRANTED. A status conference will be held on April 18, 1998 at 10:00 A.M.. At that conference, the parties will discuss the following: 1. discovery needed for trial and to determine more precisely the parameters of the class; 2. a trial date and trial procedures; and, 3. other issues relating to class certification.

SO ORDERED.