AN EMPIRICAL ASSESSMENT OF THE IN EMPLOYMENT

AGE DISCRIMINATION MICHAEL

SCHUSTER

and CHRISTOPHER

ACT

S. MILLER*

This paper presents the resultsof a content analysis of 153 federal court cases in which complainants charged a violation of the Age Discriminationin EmploymentAct. The resultsshow that57 percentof the cases were brought on behalf of white men in professional and managerialoccupations; employerswon nearlytwo of everythreecases; women were more successfulplaintiffsthan men; mor-ecases originated cases in the South than in any other region; and government-initiated succeeded more oftenthan those initiatedby individual plaintiffs.

CONGRESS

enactedtheAge Discrimina- anismforprotectingthe employmentrights

tion in Employment Act of 1967 (ADEA) "to prohibit discrimination in employmenton accountof age in such mnattersas hiring,job retention,compensation, and othertermsand conditionsof employment."' The Act and its later amendments prohibit discrimination by employers, employment agencies, and labor unions against workersages 40 to 70,2 and it has become the nation'smostimportantmech-

*MichaelSchusteris AssociateProfessorat theSchool of Managemenit,Syracuse University.Christopher Miller is a law clerk at the U.S. Court of Appeals, Fourth Circuit, Abingdon, Virginia. The Amer-ican Associationof Retired Persons-Andrus Founldation provided financialsupport for this research. 'Senate ReportNo. 95-493, 95th Cong., 2d sess., reprintedin U.S. CodeCongressional and Administrative News(1978), p. 976. 2Age Discriminationin Ernployment Act,29 U.S.C. ?621-34 (1970 & Supp. IV 1974), amendedby92 Stat. 189 ( 1978).

of older workers. To assess the record of that statute,this study applied the technique of content analysisto 153 federalcourt cases brought pursuant to the ADEA. Answers were sought to the followingquestions. What personnel actions are most likely to give riseto an ADEA claim?How successfulhave employees been in ADEA litigation?What factorshave inost influencedthe courts in deciding ADEA cases? What are the personal characteristicsof the workers who And finally, engage in thiskindof litigation? what industriesand geographic regions of the countryattractthe most ADEA litigation? In addition, the study assesses the linkage of those several factorsto the outcome of ADEA cases. Background Three majortheorieshave (lominatedthe economics literatureon employmentdis-

Industrialand LaborRelationsReview,Vol. 38, No. 1 (October 1984). ? 1984 by Cornell University. 0019-7939/84/3801$01.00

64

Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve, and extend access to Industrial and Labor Relations Review www.jstor.org

®

AGE DISCRIMINATION

IN EMPLOYMENT ACT

crimination: statistical discrimination, monopoly power,4 and personal prejudice.5These theorieshave been used extensivelyin the search for the forces behind race and sex discrimination,6 but theyhave been seldom ifever applied to employment discriminationagainst the older worker. The failure of economiststo address this issue has been somewhatcompensated for by legal scholars and industrialgerontologists,who have come to theconsensusthat age discrimination at the workplace is essentiallya manifestationof beliefscommon throughoutour society.7In particular, the pervasive belief that all abilities decline withage has fosteredthe mythin the modern organizationthatolder workers are notas efficientas youngerworkers,8 a myththat has no basis in fact.9 The recognitionthatage discrimination in employmentresultsfrom misinformed beliefs about older workers' capabilities 3See Dennis J. Aignerand Glen G. Cain, "Statistical Theories of Discriminationin Labor Markets,"Industrialand LaborRelationsReview,Vol. 30, No. 2 (January 1977), pp. 175-87. See also A. Michael Spence, "Job MarketSignaling,"Quarterly JournalofEconomics, Vol. 87, No. 3 (August 1973), pp. 355-74; and Edmund S. Phelps, "The StatisticalTheory of Racismand Sexism," AmericanEconomicReview,Vol. 62, No. 4 (September 1972), pp. 659-61. 4GlenG. Cain, "The Challenge of SegmentedLabor Market Theories to Orthodox Theory: A Survey," JournalofEconomicLiterature, Vol. 14, No. 4 (December 1976), pp. 1215-57; and Lester C. Thurow, Povertyand Discrimination (Washington,D.C.: Brookings Institution,1969). 5GaryS. Becker, TheEconomicsofDiscrimination, 2d ed. (Chicago, Ill.: Universityof Chicago Press, 1971). 6See generallyOrley Ashenfelterand Albert Rees, eds., Discrimination in Labor Markets(Princeton,N.J.: PrincetonUniversityPress, 1973); and Ray Marshall, "The Economics of Racial Discrimination:A Survey," JournalofEconomicLiterature, Vol. 12, No. 3 (September 1974), pp. 849-71. 7See U.S. Departmentof Labor, The OlderAmerican Worker: AgeDiscrimination in Employment (Washington, D.C.: GPO, 1965), pp. 51-52; and AlfredBlumrosen, "Interpretingthe ADEA: Intentor Impact,"in Monte B. Lake, ed., Age Discrimination in Employment Act (Washington, D.C.: Equal Employment Advising Council, 1982), pp. 68-115. 8WilliamL. Kendig,AgeDiscrimination inEmployment (New York: AMACOM, 1978), p. 10; and Alexander Comfort,"Age Prejudice in America," Social Policy, Vol. 7, No. 3 (November 1976), pp. 3-8. 9MatildaW. Rileyand Anne Foner,Agingand Society,Vol. 1 (New York: Russell Sage, 1968), pp. 42637.

65

mightserve well as the foundationof the formalmodel of age discriminationwe now lack. Indeed, age discriminationfitswell within the personal-prejudice theory of employment discrimination;the attitudinal cause of thisdiscrimination,for example, is an illustration of Boulding's third cause of personal prejudice: false generalizations. Other reasons given for workplace discriminationbased on age center on the higher costs of employingolder workers, such as the expense of providing fringe benefitsor trainingto these workers." Yet personnel decisions affectingolder workers are often based less on such objective coststhanon the unprovenassumptionthat older workershindermanagerialeffortsto improve productivity.'2 One obvious remedy for such attitudes is education, and indeed the ADEA provides for an education and research program.' The explosionin ADEA complaints in recent years would indicate, however, that these programs have yet to produce much change in attitudes.'4 In fact,changing attitudesabout aging may be particularly difficult to accomplish, since age discriminationis oftenviewed as a benign "0KennethBoulding, "Toward a Theory of Discrimination," in Phyllis Wallace, ed., Equal Employment Opportunity and theAT&T Case (Cambridge,Mass.: MIT Press, 1976), pp. 9-15. "See Sally Coberly, "Incentives for Hiring Older Workers-Are EmployersInterested?"Agingand Work, Vol. 6, No. 1 (Winter1983), pp. 3.7-47; A. E. Billings, "Age-related Employment Costs at the Travelers Companies in 1981," Agingand Work,Vol. 6, No. 1 (Winter 1983), pp. 7-14; and "The Cost of Growing Old: Business Necessityand the Age Discrimination in EmploymentAct," note, Yale Law Journal,Vol. 88, No. 3 (January1979), pp. 565-95. 12Employers certainlyconsidersuch factorsas fringe benefitcosts,but usually only aftertheyperceive the older worker as increasinglyless productive. This threshold decision is a qualitativejudgment easily affectedby the corporate value system.See Kendig, Age Discrimination, pp. 10-11. For examples of how suchjudgments can be discriminatory, see Buchholzv. Symons Manufacturing Co., 445 F. Supp. 706 (E.D. Wis. 1978); Scofieldv. Boltsand BoltsRetail Stores,Inc., 20 PracticesDecisions12,299 (S.D. N.Y. 1979); Employment v. SandiaCorp.,18 E.P.D. 5527 (D. N.M. 1978). Mistretta '329 U.S.C. 622 (a). ""EEOC Reports Increase in Age Discrimination Complaints,"note,Agingand Work,Vol. 4, No. 4 (Fall 1981), pp. 281-82.

66

INDUSTRIAL AND LABOR RELATIONS REVIEW

accommodationof a natural process,without the invidious intent commonly acknowledged as characteristicof race and sex discrimination.'5 It is verylikelythatvalues change more quicklyas the cost of maintainingthe old values increases,and an employer'scost of defending against ADEA complaints can be veryhigh.ADEA litigationhas, in effect, therefore,served to educate as well as to enforce. Given complainants' recent reliance on litigationrather than conciliation in employmentdiscriminationcases,'6 this propositiontakeson particularsignificance. The effectiveenforcementof the ADEA is thereforecentral to the eliminationof age discriminationin employment. The remainder of this paper assesses whether the federal courts,the Equal Employment Opportunity Commission (EEOC), and privatelitigantsare fulfilling the objectives of the Act. Methodology The traditionalmethodof evaluatingthe implementationof a statute has been to analyze the logic and consistencyof court opinions interpreting the statute. That method has several strengths,but it often fails to exploit all the importantinformationavailable in courtdecisions.This study complements conventional studies of the ADEA by applying content analysis to a large numberof the courtdecisionsinvolving that law. In content analysis,trained readers translatescatteredfactsand qualitativeinformationin primarydocuments into categoriesamenable to statisticalanalysis;17 the techniquehas recentlybeen used in another studyof employmentdiscrimination cases.'8 "5SeeBlumrosen,"InterpretingtheADEA," p. 104; and U.S. Departmentof Labor, "The Older American Worker,"pp. 5-6. '6Arthur B. Smith, Jr., "The Law and Equal EmploymentOpportunity:What's Past Should Not Be Prologue," Industrialand Labor RelationsReview, Vol. 33, No. 4 (July1980), pp. 493-95. '701e R. Holsti,Content Analysis fortheSocial Sciences and Humanities(Reading, Mass.: Addison-Wesley, 1969). '8HubertS. Feild and WilliamH. Holley,"The Relationship of PerformanceAppraisal System Characteristics to Verdicts in Selected Employment

To identifyappropriateADEA cases, we searched the LEXIS computer listing of federalcourtcases, using broad descriptive words to retrievethe largestpossible number of relevant cases. The search yielded 1,556citationsto ADEA cases heard in federal courts from 1968 through 1981. We refinedthat population in two ways. First, we excluded slip opinions (those not yet published in federal court reporters)and all cases in which another statutewas the basis forthe principalissue in thecase. Second, we recognized that every case could have more than one LEXIS citation. For cases heard in more than one forum,we recorded only the higher court's description of the case. These exclusions reduced the population to 1,151 cases. The LEXIS systemlistscases in no particularorder. The cases were read in order as theyappeared on the system'scomputer printout.Of the 1,151 cases, 612 had been read when this study was prepared. Of these, 459 were decided on procedural issues and 153 on substantive matters. These latter 153 cases form the focus of this paper. To analyze the cases, we developed a coding formthatdrew upon our previous studies of the ADEA.'9 Criteria used for selectingvariablesto appear on the coding form were their theoreticalimportance, theirpresence in at least mostof the court opinions,and theiramenabilityto statistical analysis.Those variables selected fell into one of threecategories: the personal characteristicsof thecomplainantand the organizational characteristicsof the employer, DiscriminationCases," Academy ofManagement Journal, Vol. 25, No. 2 (June 1982) pp. 392-406. For one of threeearlierstudiesin whichFred Kortemployedthis technique, see his "PredictingSupreme Court DecisionsMathematically: QuantitativeAnalysisoftheRight to Counsel Cases," AmericanPoliticalScienceReview, Vol. 51, No. 1 (March 1957), pp. 1-12. "9MichaelSchuster,"AnalyzingAge Discrimination Act Cases: The Developmentof a Methodology,"Law andPolicyQuarterly, Vol. 4, No. 3 (July1982), pp. 33951; and Michael Schusterand ChristopherS. Miller, "Performance Evaluations as Evidence in ADEA Cases," EmployeeRelationsLaw Journal,Vol. 6, No. 4 (Spring 1981), pp. 561-83. 0See, for example, Feild and Holley, "The Relationship of Performance Appraisal System Characteristics."

AGE DISCRIMINATION the case process, and the case determination. Each formprovided space to code up to eight complainantsin a singlecase, and also to code class actions.2' Eight personal characteristics of the complainant were coded: gender, race, religion,age, occupation, union membership, the structureof the employingorganization, and industry.The case reader's abilityto identifyand classifyeach variable depended on the detail provided in the court opinion and to some extent on the importanceof the issue in the case. In most cases, for example, the opinion noted the complainants' ages but not their gender. Coders inferredgender by examining the firstnames and pronouns appearing in the court's opinion. When in doubt, however, the readers alwayscoded "unknown." The complainant's occupation was divided into four categories: professional/ managerial,blue-collar,clerical,and retail/ sales. The employer'sorganizationalstructure comprised four categories: familyor individuallyowned, corporate enterprise, subsidiaryof a largercorporation,22 or government.23Industrieswere classifiedin six categories: public sector, manufacturing, 4 utilities/transportation, food/agriculture, and retail. service, The process section of the coding form categorized three items: geographic location, the initiatorof the court action, and the decision date and court of last resolution. The distinctionsin the geographic locationvariablecoincided withtheboundaries of the federalcircuitcourtsof appeals, which at the time of the studydivided the countryinto ten regions plus the District of Columbia. The variable forthe initiator of the actions classified suits brought by individuals, the government,or unions. 2'For thosecases involvinga class action,we omitted personal characteristicsfromconsideration. 22Thedistinction between"corporateenterprise"and "subsidaryof a larger corporation"was used to differentiatebetweenlarge and medium-sizedcorporate oreanizations. '3This section was later expanded to include employmentagencies and labor unions, but none of the cases analyzed in this studyinvolvedeither type of organization. 24"Food/agricultuire" denotes any organization involvedin the productionor processingof foodstuffs.

IN EMPLOYMENT

ACT

67

This variablehelped to determinewhether individualsbringingtheirown actionswere more successfulthan those representedby the Secretaryof Labor or the EEOC.25 The decision date permittedthe detection of change over time. The case-determinationsection of the coding form investigatedthree variables: the principal issue in the case, the critical factor used by the court in deciding the case, and the partythatwon the case. The principal issue in the case was defined as the personnel action prompting the controversy,such as hiring, hours of work, training,or retirement.These categories were subdivided formore precise analysis. The critical factor was defined as the aspect of the case thatthe courtconsidered the most importantin making its determination, more specifically, one of the following: 1. performanceappraisal (whether the courtagreed or disagreed withthe employer's appraisal of the complainant's performance); 2. discipline (whether the court determined that the employer's disciplinary action was reasonable and justified, or merelya pretextfor age discrimination); 3. job elimination (whether the complainant'sjob was eliminated legally,that is, because of business necessity); 4. retirementplan (whetherbona fideor illegal); 5. corporate employment policy (discriminatoryor nondiscriminatory); 6. bona fide occupational qualification or nondiscriminatory);27 (discriminatory 25"Reorganization Plan No. 1 of 1978," 43 Fed. Reg. 19,807 (1978). 26"Corporate emnploym-ient policy" refers to an organizationwideinclinationto discriminatebased on age. While this inclinationcan at times be shown by such policies as involuntaryretirement,the reverseis not necessarilythe case; thatis, an involuntary-retirement policycan be in place withouta corporatewide policyto discriminate.Thus, in one instance,the corporate employmentpolicywas the true cause of the was merely retirement discrimination, and involuntary a device to implementthe policy.In anotherinstance, the matterof the involuntaryretirementwas the discriminatoryconduct. 27Bonafideoccupationalqualifications(BFOQ) represent an affirmativedefense under the ADEA and thereforecan be a criticalfactorin any case in which

68

INDUSTRIAL AND LABOR RELATIONS REVIEW

7. medical evidence (upheld or denied); or 8. other. Next, each case was read by two thirdyear law students, who coded the cases independently. The law students compared theiranalysesand noted differences of opinion amongst themselvesand with the help of one of the authors. A third reader, an attorney,read disputed cases. Coding disparities were resolved during discussion sessions withthe authors. High levels of agreement resulted, probably because of the common training of the nature of readers and the straightforward the variables.28 Finally,because of the categoricalnature of the variablesexamined in thisstudy,chisquare testsWithassociated levels of probabilitywere performedas an appropriate method of data analysis.Nevertheless,the size of the case population and the number of categoriesinvolvedyielded an expected value in severalcategoriesof less than five. As a result, several categories were combined. For example, types of personnel actionswere redefinedas job status(including hiring, promotion, demotion, and transfer); discharge; and involuntary retirement. To facilitateand strengthenthe regional analysis,the federal circuitcourts of appeals were grouped as follows: the First, Second, Third, and District of Columbia CircuitsintotheEast; the Fourth and FifthCircuitsintotheSouth;29the Sixth and Seventh into the Midwest; and the Eighth,Ninth,and Tenth into the West.

Table 1. Selected Characteristicsof Complainants and Their Employersin 153 Age DiscriminationCourt Cases. Characteristic Complainant Gender Male Female

Numberd Percentb

104 25

81 19

Age 40-49 50-59 60-70

21 63 31

18 55 27

Occupation Professional/Managerial Blue-Collar Clerical Retail

74 32 16 7

57 25 12 5

Complainant's Employer Organizational Structure Family/Individually Owned Corporation Corporate Subsidiary Government

1 102 14 32

1 69 9 22

Industry Public Sector Manufacturing Utilities/Transportation Service Food/Agriculture Construction Retail

33 52 27 20 9 1 6

22 35 18 13 6 1 4

and aThe numberofcasesvariesamongcharacteristics, nevertotals153,becauseof missing datain thecases. bThepercentages maynotsumto a hundreddue to rounding.

ADEA (80.6 percent). Also, a smaller majorityof thecases are broughtbyprofesResults and Analysis sional or managerial employees (57.4 percent). It may be that the ADEA serves as As Table 1 indicates,men have filed a theonlyrecourseformostseniorwhitemale large majority of the claims under the professionals and managers who believe theyarean issue.Thejob action(orpersonnel action) they have unfairly suffered an adverse beingprotesteddoes notimplya BFOQ but rather employmentaction. establishes thecircumstances underwhicha claimof This propositionis strengthenedby the BFOQ becomesimportant (forexample,promotion). 28Incontrast tootherformsofcontent this analysis, lack of evidence thatwould indicatea Title analysisraisedfewissuesofjudgment.Forexample, VII-ADEA interaction.Most complainants the readerscompletely agreedon whichpartywon thevariouscases.Anoccasionaldifference ofopinion in ADEA litigation,that is, apparentlyare aroseoverthecomplainant's theprincipal not members of the classes protected by occupation, issuein thecase,and thecriticalfactorin thecourt's Title VII; not onlywere mostcomplainants adjudication of thecase. 9Partof theFifthCircuithas recently beendesig- men, but the race of the complainant was noted in onlyone case, suggestingthatmost natedthe EleventhCircuit,but thisstudydoes not includeanycasesfromthatcourt. were white.Moreover,in only21 cases were

AGE DISCRIMINATION the plaintiffsdesignated as union members, and in no case was the plaintiff'sreligion mentioned.Thus, alternativeavenues of redress, such as a union grievanceprocedure or a Title VII suit,appear to have been limited for the majorityof ADEA complainants. Title VII procedures,whichcover cases of sexual, racial, and religiousdiscrimination, are more clearly established than ADEA procedures and thus are less likely to result in procedural delays and defeats of the sortexperienced by complainantsin ADEA litigation.Moreover,discrimination based on sex, race,or religionhas generally been perceivedas more invidiousin nature than age discrimination,perhaps causing the federal courts to be more sensitiveto those formsof discrimination,30 and thus more efficientand effectivein processing those kinds of cases. Three factors may explain the apparentlylimited number of cases brought by labor union members. The firstfactor is the availabilityof, and unions' preference for, the contractual grievance procedure forresolvingworkplacedisputes,reducing their need to resort to the court system. Arbitrationcases are likelyto be adjudicated more quicklythancourtproceedings, and complainants(or grievants)in discrimination cases have at least an equal chance of success in grievance arbitrationas in court."3 The second factor is the widespread inclusion in collective bargaining agreements of seniorityprovisions,which probably provide greater protection for older workersthan thatnormallyfound in nonunion settings.Finally,the courts are generallyless willingto adjudicate a claim of employmentdiscriminationthathas not yet been processed through available administrativeor contractualremedies.) Table 1 also showsthat55 percentof the actions analyzed were filed by employees 30Blumrosen,"Interpretingthe ADEA," p. 104. 3"MargaretOppenheimer and Helen LaVan, "The Use of Arbitrationand Litigationin Employm-ient DiscriminationDisputes: An EmpiricalComparison," in IndustrialRelationsResearch Association,Proceedings of theThirty-First AnnualMeeting,August29-31, 1978, Chicago(Madison, Wis.: IRRA,1979), pp. 291-98. 32See, forexample, Alexander v. Gardner-Denver Co., 415 U.S. 36 (1974).

IN EMPLOYMENT ACT

69

between the ages of 50 and 59. That fact suggests that the Act is being used most extensivelyby those employees likelyto be in greatest need of its protection: older workerswho have reached the end of their career path witha particularorganization, who are paid more than youngerworkers, who would findit difficultto startover in a newjob, and who are not yetold enough to qualify for retirement benefits. The importanceof thisgroup is whatone might expect in lightof researchfindingsthatthe 55-and-overage group comprisesthe highest proportion of discouraged workersof any age group,33and thatthe riskof longtermunemploymentincreasessignificantly for nonworkingmen when theyreach the age of 50. 3 Since most ADEA cases are brought by male professional and managerial employees, it is not surprising that most defendantswere corporationsor corporate subsidiariesratherthan family-or individuallyowned businesses.Among industries, the public sector agencies (22.3 percent) and manufacturing concerns (35.1 perThe cent)attractedmostof ADEA litigation. former figure may surprise some who would assume that government has a responsibilityto adhere to the laws it is charged withenforcing. The resultsof the processanalysis,shown in Table 2, indicate that the largest proportionof cases took place in the South (32 percent),combiningwiththe West (24 percent)to representwelloverhalfof all ADEA litigation.These figuresmayin part reflect the lower union penetrationof the labor marketin these regions,particularlyin the statesof the FifthCircuitwhere 23 percent of all cases are initiated.) Again, this low 33Marc Rosenblum, "The Last Push: From Discouraged Workerto InvoluntaryRetirement,"IndustrialGerontologist, Vol. 2, No. I (Winter1975), pp. 1422. "'G. Boglietti,"DiscriminationAgainstOlder Workers and the Promotionof Equality of Opportunity," Labor Review,Vol. 110, No. 4 (October biternationoal 1974), pp. 351-65. 35FOrexample, withinthe FifthCircuit,only 16.5 percent of the nonagriculturallabor force is unionized. See U.S. Departmentof Labor, Bureau of Labor HandbookofLaborStatistics Statistics, (Washington,D.C.: GPO, Decem-iber1980), p. 413.

70

INDUSTRIAL

AND LABOR RELATIONS REVIEW

Table 2. Selected Process Characteristics in 153 Age DiscriminationCases.

ADEA cases are taken beyond the district court level (63.4 percent). The process analysisresulted in one particularlyinterChlaracteristic Number Percenit estingfinding:the total number of ADEA cases litigatedhas risen sharply in recent Region years. The first11 years of the legislation 39 East 26 (1968-78) accounted for only 45 percent Midwest 28 18 of the substantivecases, while the remainSouth 50 33 ing cases (55 percent)were resolved in the West 36 24 few years 1979-81. In the early years of Suit Initiatedby: the ADEA, the federalcourtswere required Individual 73 112 to resolve many procedural issues; the 41 Government 27 increasein substantivecases in recentyears CourtofLast Resolution suggeststhatthose issues have been largely DistrictCourt 97 63 settled. Court of Appeals 54 35 It is also importantto note thatthe 1979Supreme Court 1 2 81 period represents the years following Date ofDecision the 1978 amendments to the ADEA, in 1968-78 69 45 which Congress attempted to clarifysev1979-81 87 55 eral procedural issues, such as the rightto tThe percentages maynotsumto 100due to rounding. a jury trial,requirementsfor givingnotice of the intent to sue, and prohibition of mandatory retirement before age 70. rate of unionization translates into an These changes were made to strengthen of absence grievanceproceduresor seniorthe plaintiffssubstantiveclaim.At the same ity clauses that would ordinarilyserve to time,the amendments served to publicize protect older workers. In addition, 19 of the rightsof older workers,thus probably the 20 right-to-work-law statescan be found in the South and West. The employment contributing to the increase in ADEA litigation. environment created by such laws may Additionally, the transfer of enforceencourage employers to be more aggresment forthe ADEA fromthe responsibility sive in discharging or forcingthe retirementof older workers.A finalcontributing Departmentof Labor to the EEOC in July 1979 provided the Act with a higher factormay be the absence in those regions of ADEA "referral states," that is, states nationalprofileand an enforcementagency experienced in employmentdiscrimination recognized by the EEOC as having a law claims. Later discussion touches upon the prohibitingage discriminationand those havinga stateauthority empoweredto grant actual effectsof these events. The results shown in Table 3 establish complainantsrelief.36 In the South,in 1981, that terminationof employmenthas clearly for example, only three stateswere desigbeen the primarypersonnel action precipnated as referralstates.37 itatingADEA complaints.Various formsof As might be expected, because of the discharge and involuntary retirement burdensome cost in time, money, and human resources of engaging in employ- accounted for 67.2 percent of all cases. Older workers may tolerate less severe ment litigation,the governmentfilesonly formsof age-based employmentdiscrimia restrictednumber of the totalsuits(26.8 nation,and theygenerallyseem to be willpercent),38with individuals initiatingthe ing to engage in litigationonly in cases of majority of court actions (73.2 percent). separation. Also, perhaps because of those costs, few Table 3 also indicatesthe criticalfactors the courts consider most important in 36See 29 U.S.C.?633(b). 37See 29 C.F.R. ?1626.10 (1981). 38Nevertheless,many of the government-initiated suits are on behalf of multiplecomplainantsor take the formof a class action.

39AgeDiscriminationin EmploymentAct Amendmentsof 1978, Pub. L. No. 95-256,92 Stat. 189 (1978).

AGE DISCRIMINATION Table3. Selected Characteristics Associated withCase Determination. Characteristic PrincipalIssue Job Status Discipline Discharge Compensation Benefits InvoluntaryRetirement CriticalFactor Performance Discipline Business Necessity Retirement Policy BFOQ Outcome EmployerWins Employee Wins

Number Percenit 43 2 55 3 2 47

28 1 36 2 1 31

57 3 14 36 20 12

40 2 10 25 14 9

91 52

64 36

to 100due to rounding. tThe percentages maynotsUrm

reaching their decisions. The principal determinative factorsin ADEA cases appear to have been performance appraisal (40 percent), the legitimacyof corporate policies regarding older workers,which generally involve termination or forced retirement (14 percent), the legality of eliminatingemployees'jobs (10 percent), and involuntaryretirementof employees under a retirementplan (25 percent). The finding that the conduct of performance appraisals is an importantbasis forthe courts'decision rationalein 40 percent of the cases holds particular significance. The need for well-structuredand fair performanceappraisal systemstherefore seems to be critical in fulfillingthe objectivesof the ADEA.40 Yet as previous research indicates,such appraisal systems are not commonlyfound in ADEA litigation,4' despite the high success rate of in performance appraisal cases. employers 40in Cova v. Coca-ColaBottlingCo. of St. Louis, 574 F.2d 958 (6th Cir. 1978), the plaintiffsargued unsuccessfullythat formal performanceappraisals should be required as a matterof law when performanceis an issue. The court held there was no basis for the argument in the wordingof the ADEA. 'Michael Schusterand ChristopherS. Miller,"Performance Appraisal and the Age Discriminationin Employment Act," PersonnelAdministrator, Vol. 29, No.3 (March 1984), pp. 48-58.

IN EMPLOYMENT ACT

71

Table 3 reports that on a national basis employershave been victoriousin 63 percentof theADEA actions,and in thosecases where performanceappraisal was the critical factor,employerssucceeded 64.8 percent of the time (not in the table). The rate of employersuccess in general may exemplifyGalanter's propositionthat frequentlitigatorshave the advantage over less frequent litigators.42Employing his taxonomy,we can consider the employeecomplainant the "one shot" litigatorand the employer-defendant the "repeater." Their greater bargaining power and expertise may lead the repeaters to settle weaker cases out of court and to litigate stronger cases, but without further research,we cannot be sure to what extent that is true or to what extent employers succeed in pre-litigationefforts. Cross-Tabulations and Chi-Squares Our analysisusing cross-tabulationsand chi-square tests indicates a number of significantrelationshipsamong the variables in question. Table 4 summarizes the effectsof the threecategoriesof case characteristics (personal, procedural, and determination)on the outcome of the litigation, that is, on which partysucceeded. Also reported here are relationships between the geographic variables and several selected variables, and selected relationshipsbetweenage, gender,occupation, and the principalissue. Female plaintiffs had considerably greater success in ADEA suits than their male counterparts.Women won 64 percent of all cases theyinitiated,significantly more than the 29 percent won by men in the cases theyinitiated.43 At least three factors may be responsiblefor this result. 42MarcGalanter,"Whythe'Haves' Come Out Ahead: Speculationson the Limitsof Legal Change," Law and SocietyReview,Vol. 9, No. 1 (Fall 1979), pp. 95-160. 43Because of space limitations, the percentages reported here and throughoutthe discussion on the cross-tabulation and chi-square analysis are not included in any table. These percentages represent cell frequenciesresultingfrom the cross-tabulations. The differencesamong these frequenciesdetermine thesize of theassociatedchi-squaresreportedin Table 4. The authorswillsend copies of the underlyingdata to interestedreaders upon request.

INDUSTRIAL

72

AND LABOR RELATIONS REVIEW

Table 4. Selected Factors Influencingthe Outcomes of 153 Age DiscriminationCases. Variables

Nt

Chi-Square

PersonalCharacteristics Outcome x Gender Outcome x Age Outcome x Occupation Outcome x Employer'sOrganizational Structure Outcome x Industry

120 107 115 139 123

10.2 .2 .1 1.8 4.4

ProceduralCharacteristics Outcome x Region Outcome x Suit Initiatedby Government Outcome x Decision Date

143 143 143

5.2 2.7 1.0

.1591 .0998* .3219

Determination Characteristics Outcome x Principal Issue Outcome x CriticalFactor

135 106

3.7 9.2

.1604 .0099***

Geographic Region x Region x Region x Region x Region x

129 153 122 145 113

10.5 14.9 1.5 14.4 3.7

.0148** .0214** .6733 .0253** .7236

110 110 122 109

33.4 3.0 7.3 65.5

.0001*** .2225 .0266** .0001***

Gender Occupation Suit Initiatedby Government Principal Issue CriticalFactor

General Gender x Occupation Gender x Age Gender x Principal Issue Age x Principal Issue

Probability .0014*** .9039 .9442 .6238 .2249

tDifferencesamong the sample sizes for the variablesare the resultof missingdata in the cases. *p < .10. *p < .05. **'1 < .01.

First,although none of these cases was filedas a Title VII sex discriminationclaim, the factthatwomen have been grantedthe added protectonof Title VII is undoubtedly not lost upon the courts. This added expression of legislativeconcern may lead the courts to be particularlysensitive to personnel actions affectingwomen, and thusmore likelyto decide on theirbehalf.44 Second, the majorityof cases broughtby women (52 percent) involved a job status issue (hiring, promotion, transfer, or demotion),whereas only 25 percentof the male plaintiffs raiseda job-statusissue. The courts may be less willingto intrude upon management prerogativeswhen the personnel action has major financialor pro44Blumrosen,"Interpretingthe ADEA," p. 104.

ductivity ramifications, such as the dischargeor involuntaryretirementactions at issue in 75 percentof the cases brought by men. At the same time,the rightto be hired for a job that one is qualified for is a central theme in employmentdiscriminationlaw. Moreover,female plaintiffsfell into clericaloccupations 52 percent of the time,whileonly4 percentof male plaintiffs were clericalworkers.It may be thatat the low-payinglevel of clericaljobs, companies are not as attentive to performance appraisal issues as at the professionaland managerial level, where most male complainants (88 percent) were found, or to other situationsthatcould breed claims of discrimination.Thus, theorganizationmay be less well prepared to defend actions broughtbytheformergroup of employees. Table 4 shows no apparent effectof age

AGE DISCRIMINATION or occupationon theoutcomeof cases. Nor does an employer's organizational structure or industryappear to play a role in determining outcomes. As might be expected, there was a somewhat significant effecton the outcome when the governmentsued on behalf of a complainant. With the advantage of choosing the claims most likely to succeed, and the added expertiseand stafffortakingemployment discriminationsuits to court, the federal governmentwon 47 percent of all cases, whereas individual plaintiffswon only 32 percent. It should be noted that although employees' success rate in cases involving job statuswas 48.8 percent,thatrate fellto 36.5 percentin cases of dischargeand 28.6 percentin cases of involuntaryretirement. Althoughthisdifferenceis not statistically itmaybe thatemployersare less significant, attentiveto job-status mattersthan to the more dramatic actions of discharge and retirement.Also, the federal courts' hesitance to intrude on such importantmanagementprerogativesas dismissalmay hold true in general and notjust forwomen, as noted earlier. When the critical factor in the court's decision was the legitimacyof an involuntary or early retirement,employers succeeded 81 percent of the time. But when the criticalfactorinvolveda company policynot related to retirementbut adversely affectingthe older worker,such as hiring, promotionpolicies,or schedulingof salary increases, employerswere successfulonly 40 percentof thetime.It thereforeappears thatalthoughthe courtsmaybe hesitantto intrudeupon the traditionalmanagement rightto employat will,theywillact to curtail an arbitrary,across-the-boarddiscriminatorypolicy.The courts'rationalein not second-guessing an organization's retirement decision mightalso be influencedby the fact that the consequences of those decisions are often less than dire. For example, the existence of adequate pension plans oftensoftensthe blow of forced retirement,while not presentingthe court witha choice between managementrights and an individual'srightto a livelihood. Not surprisingly,there was a significant differencein the relationshipbetween the

IN EMPLOYMENT ACT

73

ages of the plaintiffs and theprincipalissues raised in the respectivecases. While 59 percent of the job-status cases were brought by plaintiffsin the 40 to 49 age group, plaintiffsaged 50 to 59 were responsible for 77 percentof the dischargecases. Similarly,the 60-and-overage group brought 63 percentof the retirementclaims.Within the youngestage group, 81 percentof the cases involvedissues ofjob status,whereas among all cases brought by those age 60 and above, 72 percent concerned retirement. As noted earlier,the plaintiffswere generallymen,but therewas a commonvariationin age among plaintiffs of bothsexes. The analysis failed to yield a significant relationship between outcome and the decision dates of cases. With the 1978 amendmentsto theADEA and the transfer of enforcementresponsibility to the EEOC in mid-1979, we mighthave hypothesized that plaintiffswould achieve greater success from 1979 on. In fact,plaintiffshave seen theirpre-1979 rate of success (33 percent) only slightlyimproved upon since 1979 (41 percent), though not in a statisticallysignificantfashion. The legislative changes of thelate 1970s mayrequire more time to fullyexert theirinfluence. Table 4 providessome additionalinsight into the effectof region on age discrimination litigation.There are significantdifferencesamong the regionsin, forexample, the gender of the plaintiffs.A lower union presence on behalf of women, who have the advantage of the large service unions in urban areas of the East and Midwest, may forcewomen to sue more oftenin the South and West. Along with the right-towork laws, thislack of representationmay make an older woman's employmentstatus especiallytenuous in those areas. Support for this point was also evident in the analysisof occupationsand principal issues byregion.The South and Westshow considerablyhigherpercentages(38.1 and 34.6, respectively)of suitsbroughtby clerical employees than occur in the East and Midwest.If we recognize women's greater compositionof the clericallabor forcethan of the professional or managerial labor force, their strong representationamong the plaintiffranks in the South and West is not surprising.

74

INDUSTRIAL

AND LABOR RELATIONS REVIEW

the spectre of conflictbetween the objectivesof the ADEA and those of Title VII.46 The employmentrightsof older workers, that is, may be competing with those of workersprotectedunder Title VII, and the EEOC may finditselftornbetweenenforcing two pieces of major social legislation withcompetingconstituencies. Men may have been the most frequent litigantsin ADEA cases, but women were the more successful. The protection afforded women under Title VII may indeed be influencingthe treatmentand outcome of ADEA claims involvingfemale plaintiffs.In thisinstance,the goals of the ADEA and Title VII may be equally well served. For both sexes, the principal issues of dischargeand involuntaryretirementhave been responsibleforthebulk of substantive ADEA litigation.In viewof the seriousness of such personnel actions, this findingis not surprising.Among women alone, however, the majorityof cases concerned jobstatusissues. Perhaps the traditionalposition of women in low-payingclerical positionsmakes them not so much easy targets for dismissal but easier workers to manipulate. A major limitationof this study is the absence of comparable data on conciliation effortsin ADEA claims thatnever reached the federal courts. Since the vast majority Conclusions of complaints of age discrimination in The resultsof thisstudysuggestthatthe employment are resolved before ever Age DiscriminationAct has become the reaching litigation,an analysis of federal primarydevice used by whitemale profes- court cases is in no way a complete assesssionals and managers to attack arbitrary ment of the Act. Future research should personnel decisions. Although this may investigatethe nature and impact of conhave been an inevitable, if unforeseen, ciliation effortsand administrativeremedevelopment,thisuse of the Act does raise dies at both the state and federal levels.

The results also indicate that ADEA actionsin the South and West were undertakenmoreoftenthannotto redressalleged discriminatory discharges(42.5 percentand 52.9 percent,respectively).These findings furthersupport the propositionthatthese regions are less conducive to employment security. Finally,although the assessmentof case outcome by region indicatesno statistically that significantdifferences,itis noteworthy the success rate foremployersin the South (51 percent)was far below the success rate foremployersnationally(64 percent).Here, we mightposit thatthe right-to-work environmentand a low levelof unionizationpermits organizationsto be somewhat more recklessintheirconductofpersonnelmatters. At the same time, the prominent role takenby the FifthCircuitCourt of Appeals in ADEA litigation45and the proficiency the Circuit developed in the handling of many ADEA claims may have served to facilitateand refinethelegal processin that region. It may be, thatis, thatmanyof the procedural issues thatoftenconfound new legislation were overcome by experience, allowing the respective federal district courts to address the merits of disputes more appropriately.This experience may be accruing to the benefitof plaintiffs.

45See, for example, Marshall v. GoodyearTire and RubberCo., 554 F.2d 230 (5th Cir. 1977); and Cokev. GeneralAdjustment Bureau, Inc., 640 F.2d 1,226 (5th Cir. 1981).

46Blumrosen,"InterpretingtheADEA," p. 105; and Carl E. B. McKenry,"Enforcementof Age Discriminationin EmploymentLegislation,"HastingsLazvJournal, Vol. 32, No. 5 (May 1981), pp. 1157-93.