c 3|: gm“ . VII'I- Mu. 1","!lzllv'” . '1‘. .' ' ‘ *" JR“ ;I:I-((I: Irrflfafl‘ ”1&5? ‘ "' In. £1”; 1" . I f‘ , .1 .j ‘ .. ',- "ar- - . ' ' _>'.,I'T._;u.k ‘ .5 0“} I" .III . 5' 5%“: , ISI‘ #4 QWIIIIILI“ “I ”.3“ up.“ 3,74%;qu I II-\' I- ‘II .' . ,I “Id? :0” ‘ ‘:. H. (u '." fig?“ - ‘ . I AN” “’I 3 "'I ' |:‘. :I':‘I’I!I‘ lit’tnf' 'I: )u .. pl.» NO ‘: I " y . a: fig 1‘;.t.<.,,"'s.l}w~‘ (j; - ""‘ ‘1}, 3;" 1}" .Mfl.’1’r"z" f .mII IJ'JI'Vi‘Q‘h. {I‘IJL'L 9.? dl-‘Iv " I‘M If: Lt? Ill IIIII'IIIH‘ (NW/HG} m”: .35“: I.:,Lv.r:.1.l.'t'f3' " ’2 . . . .. 'mv. I'M ".- '1‘4 Ew‘ 1 $1.0.» . \"E 0‘ r‘ #!.1\.‘ IJII“ I‘I'O'Hx‘fr' TI”. J}: :fl‘IJ' 'l "Riva “W. V. ' . I' II; I IIIIII:;\..I|."‘ :. [3-0. ‘_.-;rl1;,",; In: "A I II _I sIzIII.(- 'I ’ Au," .1 J "'1‘ $‘31&-: 4113;; defl «,1; t.Iu1H}|'\'ul.n I III 5: 1’3". .-. ;; . .. .:v we: ° .3» ' -..w’- . 5:0: . - ' ' . s, "» 'n' .l.\‘." “-...” "Um-,3 : ' . .. '3‘?» In" ‘; WI , w. '_I :. .“-- .: n: IIIpri'I' -:- ' * *«-»~.:=-';‘ .4, ., .. . IIIIII'I ‘IuInhI ‘4""I‘Ifrh' :I a With Ila” 5F}. ' ‘h \If’ri'l. V4 352335;? “9" 11""??? IL&(M€? 'It'x‘ I’x'I'xvj' . ‘ \t.';“ I”? ,1 “ . u‘ Mtnm :"IM "F H). +(3'f I'In‘ . :92: . éifi o"’" I . 'Tfr It.“ I' I05". II IrIIéI - I..,I LI ".‘ .rijt r J: . I . why yI I.‘:-,.III .xIIII I'lII§I.I.IIIu!IIIII I I ,. .I:)I.:V- l"|I>‘ ' if III I "Y I Ir, -.‘ ~ ‘ . | I. ,5 '. :K“! "I -" 1“ tl‘l‘: mu" ‘1'”.‘1flu‘flg-p. i '4. . I. . 17" . I- . --3\,:III 121”“ _ ' ,. If "WNW? f "In ‘1 ’3 bob-co, u 93- u» .IIH- . ' :. |I “I" ‘1‘. I‘II‘IIxIV. II‘II.", ’ \ ‘4. "fin“ - I' -3 “I"? I. I|II_ I'1III :. H ~0|. .6“ fl 4 v 0 -ft’ ' THESYS l\\\\\°\l\'\°\\l\\llll\l\lll\ll\ll\\\\\\\\\\\\\lj\\l 1/ . 1‘ LIBRAR 3' Michigan State University I ————— This is to certify that the dissertation entitled Minority Representation in Municipal Workforces: An Assessment of the Effects of Affirmative Action presented by Lana Sharon Stein has been accepted towards fulfillment of the requirements for Ph.D. . Political Science degree in <1“th Amt Major professor Jack H. Knott Date M MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 IVIESI.J RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from “ your record. FINES will be charged if book is returned after the date stamped below. :- P v: 4% r' I v ’7 "-3 k 4:7 ’3 iA?‘ 3'“ 45' 1 29: - 1- 2 ” h~“ . OCT?) 3 ”2033’ 1 00:1174 ‘ \ ' I. " ,. _--'\a my; .21 4. w-J w. MINORITY REPRESENTATION IN MUNICIPAL WORKFORCES: AN ASSESSMENT OF THE EFFECTS OF AFFIRMATIVE ACTION BY Lana Sharon Stein A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY 1984 @Copyright by LANA SHARON STEIN 1984 ABSTRACT MINORITY REPRESENTATION IN MUNICIPAL WORKFORCES: AN ASSESSMENT OF THE EFFECTS OF AFFIRMATIVE ACTION By Lana Sharon Stein The principal premise of this research is that implementation of a federal program or administrative order in the manner intended cannot be tacitly assumed. The attitudes and actions of implementers, particularly if they are of a different level of government, i.e., municipalities, may be critical to the success or failure of the federal policy. Affirmative action policy represents a federal administrative order applied to local government in 1972. As its legitimacy has been widely disputed, it presents a rigorous opportunity to test the importance of local political and organizational factors on implementation. These relevant local factors include political form, bureaucratic autonomy and presence of a minority executive. It is hypothesized that a council- manager form of government will be less responsive to an order which benefits minority groups than will a mayor-council form. The latter has been judged in various works to be more responsive to racial and socio- economic cleavages. Second, autonomous bureaus which self-determine their rules and procedures and consider themselves professionals, i.e., formal civil service commissions, will be less responsive to a politi- cally-based order requiring procedural changes to ensure implementation. Finally, in the pluralist tradition, a minority Officeholder will provide impetus to the increased hiring of minorities. A regression model is used to measure the impact of local factors on a policy-specific dependent variable, namely percent of minorities in the workforce weighed against percent of minorities in the city population. Variables are also included representing local economic conditions, demographics and enforcement effort. The total city workforce is examined as are the protective services, police and fire, municipal units considered most resistant to integration. The results of model estimation demonstrate the salience of local political and organizational factors in the manner hypothesized when economic controls are present. The results are a good beginning toward understanding the implementation process and contain the seeds of further research to enhance the degree of variance explained. For Carolyn and Kristin Gaulden ACKNOWLEDGEMENTS The research effort detailed in the pages to follow has been made possible and profitable because of the encouragement, suggestions and well-taken criticism of a number of individuals. Of course, whatever shortcomings remain in the work are mine alone. I am indebted to my guidance and dissertation committee chair, Jack Knott, for valuable instruction, advice, critical judgment and friendship. The continuing cooperation and insight I received made the project manageable and also permitted its placement in a broader perspective. I am also very grateful to Gary J. Miller and Charles Press for their well-drawn criticism and support which has been both constant and valuable. In addition, I would like to acknowledge the importance of the methodological instruction provided by Charles W. Ostrom Jr. Critical assistance in every facet of computer operations has been provided by Harriet Dhanak. I am grateful also to the Political Science department and its chairperson, David Rohde, for assisting with the costs of data acquisition. The resources and support provided me by Political Science faculty have been augmented by a network of friendship stretching both inside and outside the political science discipline. The following friends have strengthened my confidence and enriched my enjoyment of both working and nonworking hours and deserve individual acknowledgement: Phyllis Gaulden, Mary Wren Bivins, Kevin Marken, Margaret Gilkison, Gaye Gilbert Benson, Karen Albrecht. Thank you. iv For my interest in political life and political impacts, my father, Irving Stein, was responsible. From my mother, Louisa Stein, I received the necessary encouragement to begin again. My cousins, George and Ruth warren, also gave impetus to my desire to return to school in ny thirties and pursue an option that had seemed unobtainable. George, as a career public personnel administrator, certainly stimulated my choice of topic though I daresay he might question my interpretations. Finally, I would like to acknowledge the assistance and support of Sue Plesko and Karla Bellingar, who typed various and all pieces of this document at the different stages of its development. Their speed, accuracy and encouragement greatly aided this process. TABLE OF CONTENTS Chapter I Affirmative Action and Municipal Employment: Establishing the Context . . . . . . . . . . . Implementation of an Administrative Order Affirmative Action Policy . . . . . . . . The Federal Role . .p. . . . . . . . . . . Local Factors Affecting Compliance . . . . Political Form . . . . . . . . . . . The Autonomous Professional Personnel BUtEBUeeecoeoeeoeoeoo Minority Chief Executives . . . . . . FootnoteSOOOOOO00.00.0000... Chapter II Specifying a Model of Minority Representation in Mun1Cip81 Workforces e e o o e o o o o o o 0 Earlier Models of Minority Representation Initial Dimensions of Analysis . . . . . . Initial Methods of Analysis . . . . . . . The Sample . . . . . . . . . . . . . Analysis of Variance . . . . . . . . The Model . . . . . . . . . . . . . . . . Analysis of Results . . . . . . . . . Footnotes . . . . . . . . . . . . . . . . . . . Chapter III Initial Statistical Findings Concerning Local Political and Organizational Factors . . . . . Analysis of Variance According to Key Local Dimnsions o o o e e o o e e o o POliticalFormoooooeoooeo Bureaucratic Autonomy . . . . . . . . The Presence of a Minority Exeuctive vi 10 ll 15 20 24 29 29 33 39 39 41 43 46 48 49 51 51 53 55 Key Determinants 1“ Pairs o o o o e o o c o o 57 The COIHPIEte Typology 00000000000. 62 The Mayor-Council Cities: Variations Due to Partisanship and the Nature of Council EleCtions00.000000000000000 64 Correlational Analysis 0 o o e o o e o o o o o 66 FOOtnOteseoooooooooooooocoso...73 Chapter IV Modeling Minority Workforce Representation: EXPCCtations and Realities o e e o o e e o o o o o e 74 PrOteCtive Serv1ces o o o o o o o o o e o o o 0 8‘. Modeling the Workforce without Protective serV1cesooo00000000000000...91 An Assessment of the Results . . . . . . . . . . 93 FOOtnOtesoeoooeoeoooeoooooooooo95 Chapter V Minority Municipal Workforce Participation: Toward Explaining the Unexplaigedfigfilm.w._. . . . . . 96 “Ch—-5 4 h—.—- f..- .- , Hiring Minority Workers: An Allocational Strategy“,00000000000000.000097 The Organizational Conceééé"“EESfEEEISEEIIEEWWH and Authority . . . . . , , . . . . . . . . . . 101 The Personnel Profession . . . . . . . . . 101 The Protective Services . . . . . . . . . . 104 Federal Enforcement Efforts . . . . . . . . . . 108 Economics and Demographics . . . . . . . . . . . 109 In Summation . . . . . . . . . . . . . . . . . . llO Footnotes . . . . . . . . . . . . . . . . . . . . . . 112 Appendix Sample Cities and Their Characteristics . . . . . . . 113 Bibliography 00000000000000...00000000000117 vii 2-2 2-3 3-9 3-10 3-11 LIST OF TABLES The Political Form Dimension: Receptivity to Affirmative Action The Bureaucratic Autonomy Dimension: Receptivity to Affirmative Action The Ethnic Politics Dimension: Receptivity to Affirmative Action Local Determinants of Affirmative Action Implementation Minority Representation in Central City Protective Services, 1967 Average Shortfall in the Achievement of a Representative Workforce Mean Ratio of Minority Representation Minority Workforce Representation by Political Form Minority Workforce Representation by Bureaucratic Autonomy Minority Workforce Representation by Presence of a Minority Executive Minority Workforce Representation by Political Form and Bureaucratic Autonomy ' Minority Workforce Representation by Political Form and Presence of a Minority Executive Minority Workforce Representation by Bureaucratic Autonomy and Presence of a Minority Executive Minority Workforce Representation by Political Form, Bureaucratic Autonomy and Presence of a Minority Executive Effects of Partisanship and Type of Council Elections on Minority Workforce Representation in Mayor-Council Cities Correlations between Minority Representation in the Total WOrkforce and Key Determinants Correlations between Minority Representation in the Protective Services and Key Local Determinants Correlations between Minority Workforce Representation Excluding Police and Fire and Key Local Determinants viii 35 36 37 38 4O 50 51 52 54 55 57 59 61 63 65 67 69 70 3-14 4-1 5-1 LIST OF TABLES (continued) Minority Representation and Political Form: Controlling for City Size and Minority Population Model of Minority Representation in the Total Workforce Two Models of Minority Workforce Representation: A Comparative Asses sment Model of Minority Representation in the Protective Services compared to Model of the Total Workforce Model of Minority Representation in the Workforce with Protective Services Excluded Compared to Model of the Total Workforce Correlations between Receptivity to Minority Hiring and Measures of Professionalism ix 71 78 81 86 92 103 Chapter 1 Affirmative Action and Municipal Employment: Establishing the Context Implementation of an Administrative Order Beginning in the 19703, a number of studies appeared which examined the process by which policies enacted by Congress are turned into func- tioning programs. Classical public administration had assumed that the carrying out of policy directives would be an exercise of hierarchical command and control and therefore nonproblematic.l However, the recent work in this area has shown that implementation in the manner intended cannot be assumed. After examining the failure of Economic Development Administration efforts to create jobs in Oakland, Pressman and Wildavsky stressed the idea that the organizational mechanisms needed to implement a program must be addressed within the policy design process and must be given as much thought as the strategy to be used for policy adoption.2 Van Meter and Van Horn continued the emphasis on the importance of the organ- izational mechanism and highlighted the role played by actual imple- 3 menters, "a crucial role in shaping the policy process.” To them, a key object of study must become ”the personal, political and organiza- tional forces at work within the implementation arena itself."4 This vein of analysis was continued in the work of Berman and McLaughlin on the adoption of innovational techniques in public educa- tion. They found that successful implementation occurred only if a 2 mutual adaptation took place between the project design and the institu- tional setting.5 Such adaptation was found to be dependent on both the organizational climate and the principal actors.6 For an innovation to be adopted, it had to be seen as central to a school district's priori- ties, receive general support from district officials and the active support of school principals.7 Unless such support for the project existed, individual teachers, the direct implementers, would not alter traditional patterns of behavior. "The receptiveness of the institu- tional setting to the change agent project seemed to be a necessary condition for successful implementation.8 Although these works have studied program rather than policy direc- tive implementation, it is apparent that organizational climate and the support of organizational actors will be critical to the successful implementation of administrative orders as well. In fact, several authors have remarked that the processes for implementing administrative orders and programs are part and parcel of very similar circumstances and 9 The question of organizational climate and the needs and conditions. attitudes of the implementers are constant. The implementation process is made more complex by the fact that policy directives for social regulation and various social programs are enacted at one level of government and administered by another. The administrators, local governmental actors, enjoy considerable autonomy and ”the strength of local institutions varies from city to city."10 Certainly local governmental capacity can affect the implementation process as can the attitude of local officials. In fact, a number of studies have noted the failure of federal planners, executives and 3 legislators to consider local circumstances with adverse consequences for the programs they wished implemented.ll Because of the degree of autonomy local governments enjoy the federal government is limited in the measures that it may employ to enforce compliance with its administrative orders. Nakamura and Smallwood note that it is impossible, for example, for the federal government to abolish local governments if they fail to carry out federal policy directives, or to fire local employees who defy these direc- tives. 12 Compliance can revolve around how local officials perceive the benefits of a federal order or program and the costs they feel they may incur if they do not actively implement. An analogy is drawn by Mazmanian and Sabatier. They note that, Several studies of compliance with judicial and admin- istrative decisions have demonstrated that, in practice, behavioral compliance is generally related to individuals' assessment of the relative costs and benefits to them of following legal directives. These same studies have suggested that the decision to comply is, in turn, a function of (a) the probability that noncompliance will be detected and successfully prosecuted; (b) the sanc- tions available to penalize.noncompliance; (c) target group attitudes concerning the legitimacy of the r es and; (d) the costs to target groups of compliance. Most important, the federal government lacks effective compliance mechanisms. As stated earlier local governments cannot be abolished nor can their powers be assumed by federal executives. The most potent penalty is the denial or cutoff of federal funding. Yet, this can be a tricky compliance mechanism. If the federal government withdraws funding from a city, the cost to citizens in lost benefits and services can be A high. Generally, the citizens most adversely affected by such losses would be minorities and the poor. However, if threats to withdraw funds are not carried out, the legitimacy of the threats is lost.14 In reality, the threats to cut off program funds vastly exceed the instances where a cutoff has actually been announced. Hence, it is not surprising that Bardach feels that the ability of the federal government to influence local actors by threats of funding cutoffs "is so slight that we could reasonably call it an inherent limit of this implementation mechanism."15 Though the federal role in the implementation process cannot be dismissed out of hand, it becomes apparent that the actions of the local government may be decisive. Therefore, guided by the works of Van Meter and Van Horn and Berman and McLaughlin, attention must center on the local implementation context where analytical procedures can be employed to outline the political and bureaucratic factors which affect adoption of federal programs or administrative orders. In order to isolate key factors in a local environment which may affect compliance with federal policy, a specific policy area will be examined, namely administrative orders requiring affirmative action to increase minority representation in municipal workforces. Changes in municipal workforce composition will be examined eleven years subsequent to the application of affirmative action orders to local government. Affirmative action has been chosen as the policy area for several reasons. First, it will provide expanded knowledge on the implementation of an administrative order. Second, it does necessitate action and commitment by urban officials, elected or appointed, and considerable 5 behavior modification by career civil servants who will be asked to supplant their professional standard of merit with a new politically- determined criterion. Third, it provides an assessment of the relative strength of minority chief executives to influence workforce composition in the tradition documented by pluralist writers. Finally, the order is a controversial one which has aroused opponents within and outside government. Hence, the legitimacy of the order may be questioned by those who have to implement it as well as by various groups within a local community. The implementation process could not be expected to be easy or quick, thereby allowing ample opportunity for the examination of certain key features either inhibiting or expediting compliance with this federal order. It should be noted that some cities may have expedited the hiring of minorities prior to 1972 for various reasons. However, records dealing with the racial composition of employees prior to that date generally were not maintained and, in fact, many cities did not comply in the area of recordkeeping for some years after coming under the coverage of antidiscrimination statutes and federal affirmative action orders. Affirmative Action Policy Discrimination in employment based on race, sex, religion or national origin was prohibited by Title VII of the Civil Rights Act of 1964. In the years following passage of this Act, the federal government moved beyond enforcement of equal opportunity to require recipients of federal funds ”to remove all barriers, however informal or subtle, that 6 prevent access by minorities and women" to employment.16 The U.S. Equal Employment Opportunity Commission stated that, "the most pervasive discrimination today results from normal, often unintentional and seemingly neutral practices throughout the employment process."17 In conjunction with the Supreme Court's ruling in Griggs v. Duke Power, the federal government required that job qualifications be significantly related to job performance and that nondiscriminatory standards be developed.18 Further, employers were required to pinpoint underutil- ization of women and minorities in their total workforce and within all subunits in order to set both short term and long range goals and establish timetables to overcome any underrepresentation of women and minorities, in terms of representation in the population and the total workforce of the local area. (In reality, population is most often used as the benchmark because figures are more readily obtainable.) Local governments were not affected by the provisions of either Title VII or initial affirmative action order until 1972. The Equal Employment Opportunity Act, passed in that year, subjected cities to the same anti-discrimination requirements already affecting other employers. Compliance with affirmative action directives would require concerted effort on the part of many local governments. The U.S. Civil Rights Commission had documented the underrepresentation of minorities in municipal governments.19 As Rodgers and Bullock stated, local governments practice many of the same discrimina- tory acts found in private industry. They 1) discrimi- nate in hiring and promoting minority group members; 2) make no special effort to recruit minority group members; 3) use methods of hiring and promoibng which place minorities at a serious disadvantage. 7 Thus, in order to undertake affirmative action, most local governments had to alter fundamentally recruitment and selection procedures. Whether change would take place in a timely manner is problematic because of the federal role in this policy area, the policy's controver- sial nature, past practices of local government and the extensive behavior modification needed for compliance. Many cities have also faced financial constraints in the late seventies and early eighties which have affected hiring and job security. It should be noted at this time that it is possible that hiring to fill already available slots may provide an inexpensive way to begin to answer the many demands made of local govern- ments by minorities from the 19608 onward.. If hiring were to be viewed in this light, this might mitigate various factors seen as obstructing compliance. The Federal Role Federal affirmative action orders represent, to Bardach, an example of ”piling on,” an accretion to an original doctrine which may change its nature and undermine its original supportive coalition.21 He states that, the concept of ”affirmative action” in the United States once meant a commitment to give preferential treatment to job applicants from racial minorities when in other respects they were "equal” to applicants from nonminority backgrounds. Over the last ten or fifteen years, the concept has come to mean quotas and a d Emphasis on the ”equality-in-other respects” criterion. Affirmative action came to mean not just creating a place for equally- qualified minorities; rather, it necessitated an examination of and 8 changes to employment selection patterns of long duration. As such, its prospects for rapid implementation became less favorable. More redrafting of behavior patterns was needed and some traditional support among white groups diminished.23 The augmented affirmative action directive, requiring additional effort, was not initially conveyed clearly to the nation's cities. According to Miller, cities were confronted by a number of federal agencies requiring affirmative action in employment. Each ”stresses different aspects, requires different reporting forms, and gives different advice"25 and covered various portions of the workforce. Special grants, such as Community Development Block Grants, contained additional affirmative action provisions, namely the hiring of those of low income and residents of project areas.26 At least initially delays could be expected in the implementation of affirmative action for municipal personnel selection because of lack of clarity regarding what was to be done. Further delays might be the result of lack of desire to change. In any case, the various federal agencies which monitor affirmative action might wish to act to speed compliance. Two routes were open to them: funding cutoffs or court suits. Ideally a funding cutoff for lack of compliance with affirmative. action could be a potent compliance mechanism. The federal government could deny funds for a wide range of programs rather than for just one program area.27 However, Bardach and others mention the severe limitations to this mechanism and, in fact, inquiries to the Civil Rights Division of the U.S. Department of Justice showed that no city has ever 9 had its federal funding terminated for lack of compliance in this policy area though there have been threats and publicity regarding such threats. It is possible for the federal government to enter court suits on behalf of private litigants who charge employment discrimination. The suits are handled by the 0.8. Equal Employment Opportunity Commission (EEOC) or its designated state agency (for example, the state civil rights commission). But, suits against local government must be prosecuted by the 0.8. Attorney General. Court suits may take a consid~ erable period of time. Whether prosecution takes place at all is prob- lematic. Herbert Hill has documented that the EEOC initially did not attach a priority to public sector complaints and had accumulated a back— log of over 100,000 local government cases.28 Further, it must be noted that the EEOC has been primarily a passive administrator, plagued by insufficient and poorly trained staff and high employee turnover.29 Its success rate in complaint investigation has been poor, with only 6% of its cases successfully resolved in its first seven years of operation.30 During the Carter administration, the EEOC was given greater juris- diction. At this time, it did reorganize and create new procedures to process complaints more quickly.31 However, the Reagan administra- tion has not continued a commitment to affirmative action and has put forth guidelines which reduce the government's role.32 A clear demonstration is the position enunciated in early 1984 by a 0.8. Civil Rights Commission dominated by Reagan appointees; they, in effect, declared that affirmative action was an inappropriate remedy for whatever employment inequities might exist. 10 Thus, it appears that the federal government's efforts at compliance may be neither constant nor certain. Hence, whether affirmative action is successfully implemented or not at the local level will likely be principally a function of the disposition and actions of local imple- menters. If such implementers are not committed to the policy objective, significantly modified behavior is unlikely.33 Commitment to the policy objective, or perhaps in this case, less resistance, can be based in part on certain local political and organizational factors. Local Factors Affecting Compliance In assessing the effectiveness of affirmative action in rectifying minority underrepresentation in municipal workforces, federal enforcement activity and socioeconomic factors will definitely affect the hiring of minorities as well may the population composition of each city. However, it is also strongly felt that three local political and organizational elements may expedite or impede compliance with affirmative action. These same elements may come into play in the allocation of municipal resources to serve various groups and in the implementation of a number of redistributive programs and are as follows: 1. Political Form: Can the structure of local government he a deterrent or an aid to implementation? Does a mayor-council form provide a more open vehicle for the address of interests of minority groups (such as affirmative action) than the council- manager form? 2. Autonomous Professional Bureaus: Will a local bureau which has protected itself from outside interference in the determination ll of its norms and procedures, such as a civil service commission, resist the change required to its professional conception of merit, and the activities derived from that conception, by federal affirmative action orders? Does it have the ability to resist? 3. Ethnic Politics: In the tradition documented by pluralist writers, will the presence of minorities as elected mayors or city managers further the representation of minority workers on the municipal payroll? Political Form Jeffrey L. Pressman has stated that ”the quality of political leadership and political institutions at the local level” are ”factors which can have an effect on the implementation of federal pro- grams.”34 More emphatically he feels that ”the achievement of federal urban program goals is dependent upon the existence of strong local political leaders who can offer support for the programs."35 Whether there are strong supportive local leaders in the case of affirmative action is partially a function of attitude, ideology, and political base, i.e., whether there is inherent support for the goals of the policy. However, it may also be strongly dependent on the form of government a locality possesses. Cities may be divided into two major types: council-manager and mayor-council. There is considerable varia- tion within these types but the variation between them is notable. As Adrian and Press state, The forms of government . . . affect the pattern of influence of various groups upon policy making. The specific structure in any given case helps to establish behavior patterns and attitudes toward power and the exercise of power that defigitely affects the process whereby decisions are made. 12 The council-manager form of government was designed to create professional, depoliticized administration. As urban chief executive, the city manager ”was expected to stay out of politics” and was expected ”to be motivated by professional, rather than political, incen- tives."37 Various urban scholars have commented that this form--small, part- time nonpartisan council members elected at large and professional, expert manager as administrator-does not provide a forum reflective of the various socioeconomic and racial differences in the locality.38 In fact, Banfield and Wilson have noted that original supporters of council-manager government wanted ”the satisfaction of putting handicaps in the way of the 'great unwashed' on the other side of the railroad tracks."39 For reformers, ”making government 'businesslike' meant 'getting rid of politics,‘ which, in turn, meant curtailing the represen- tation of low-status minorities.”40 Today, council-manager governments are still seen as often "very sluggish in their reaponsiveness to the demands of emerging political forces” and may also be viewed ”more as an impediment than as a facili- tator of 'good government' by those demanding participation and change.”41 As affirmative action is designed to be of benefit to minorities, it is conceivable that it would fare less well under a struc- ture of government which "weakens the representation of particular groups, especially if these groups are of lower status."42 The council-manager cities themselves are more likely to have a relatively homogeneous population with a large, stable middle class. Also, in keeping with their depoliticized nature, they tend to have few 13 organized interest groups representing coalitions of citizens.43 Jeffrey L. Pressman said of one council-manager city without active parties or groups that ”it often appears as though politics in Oakland does not exist."44 Lineberry and Sharkansky report that a survey of city managers has shown that the managers have a ”basic professional orientation toward efficient management and delivery of public services and towards detailed financial accounting and personnel administration. Their skills are administrative and their major resource is technical expertise.“45 They are less inclined to "shoulder the responsibility for the formula- tion of social policy.”46 Conceivably, they would be less inclined to bear the burden of social regulation, such as affirmative action, especially as the local political form does not appear conducive to the redress of grievances based on racial or socioeconomic cleavages. Of course, there will be variation among council-manager cities as there will among mayor-council cities. The latter group can vary greatly within itself as to form: strong mayor/weak mayor, partisan/nonpartisan elections, district/at large council selection. Adrian and Press note that most mayor-council cities ”represent a compromise between the very weak and the very strong mayor plan.”47 However, they feel that the mayor-council form can allow for "firm leadership” and does allow for the "placement of administrative problems in a political context."48 In addition, mayor-council cities are likely to be more open to ethnic, neighborhood, and other special interest groups.49 In his/her dependence on a coalition of groups to achieve the office of principal administrator, an elected mayor may prove to be more attuned 14 to affirmative action's natural constituency, minority citizens. Further, the mayor-council form of government appears to allow better for the expression of demands from various social groups. Some empirical evidence for this proposition can be found in the work of Lineberry and Fowler. They hypothesized that "socioeconomic cleavages will have less impact on the political choices of reformed than unreformed govern- ments."50 Indeed, they found that with reference to taxation and expenditures policy ”reformed cities are less responsive to cleavages in their population than unreformed cities."51 Lineberry and Fowler conclude that the public policy of reformed, or council-manager, cities "takes less account of the differences between White and Negro, business and labor, Pole and WASP."52 They end by saying that, The translation of social conflicts into public policy and the responsiveness of political systems to class, racial and religious cleavage differs markedly with the kind of political structure. Thus, political institu- tions seem to play an important role in the political process-a g le substantively independent of a city's demography. Given this literature, it is likely that a city's political form may affect its compliance with federal affirmative action orders. Of course, some city managers may implement because it is the law or they wish to avoid tangles with federal agencies. However, it seems the possibility of greater responsiveness to minority demands exists in a mayor-council format. The two structural types appear to present a tradeoff between centralization with limited representation and responsiveness and more fragmented government with greater citizen access.54 How responsive either form is, especially the mayor-council, is also dependent on the 15 autonomy of a city's personnel bureau and whether the chief elected or appointed official is able to provide direction to that bureau. The Autonomous Professional Personnel Bureau Students of public organizations have debated the question of whether a bureau could be both efficient and accountable to elected leaders and the general public. Using efficiency to indicate expeditious performance within prescribed cost constraints, one finds that a number of agencies do qualify as efficient. However, such agencies may often also enjoy a de facto independence from political or popular censure. The accretion of a base of autonomous power and/or jurisdiction permits the exercise of functions or delivery of programs in a protected and self-determined manner. The various agencies Robert Moses controlled in New York are clear-cut examples of the phenomenon. However, within city administration itself, departments can be identified which are efficient by their own definition and also impervious, to one degree or another, to political direction. The growth of autonomous, professionalized urban bureaus was made possible by the weakening or elimination of a politically-based local executive during the Progressive reform period. Though both elected and unelected officials have attempted to control these bureaus, many have resisted the directives of such officials. Instead, the bureaus sought to determine policy in terms of their own professional norms. In describing New York City, Lowi maintained that it has become almost impossible for a newly elected mayor to establish "coordinated 16 control of these virtually autonomous agencies."55 Similarly, Sayre and Kaufman found that in New York these ostensibly nonpolitical, ”expert” bureaus assert ”that the Mayor and other 'politicians' (even though they are elected officials) should not 'interfere' in police administration, hospital administration, park administration, or other 56 The existence of such bureaus can domains of city government.” affect delivery of federally-funded programs in law enforcement, community health, community development and other policy areas. 0f relevance to the question of affirmative action is the fact that the first of these professionalized bureaus was the civil service. Civil service was instituted to replace political hiring with selection by merit. As the determiners of merit, civil service staff were quickly able to achieve organizational separation and semi-independence. This independence and separation "provided encouragement to the development of 57 Civil scientific techniques in the personnel field itself." service staff considered themselves to be professionals applying these scientifically-based techniques and principles‘to employee selection. Examinations to "measure . . . qualifications objectively and competi- tively"58 came into widespread use as did position classification. ”Jobs could be studied in terms of the duties involved and the qualifica- tions necessary to execute them; they would be systematically differen- tiated one from the other and standardized into classes having similar requirements."59 As these particular professional practices became standard in public personnel bureaus, their practitioners began to seek "separation and autonomy” as ”protection from politicians, administrators, employees or 17 interest groups."60 They brooked no interference in their profes- sional decision making. Rather, they alone would determine how to hire according to merit. Sometimes, autonomy was assured by placing a civil service commission and its director beyond the purview of elected officials; commission members or the director were allowed to serve indefinitely and could only be removed after a hearing for cause. As Mosher concludes, during the reform era, civil service administration became a respected specialty, requiring focused educational background . . . . It was a near-science, an emergent profession. It treasured its techniques, its objectivity, its impersonality, its neutrality. It was not a tool of political party or pressure, of legislators or administrators. It was, or aspired to beche, a triumph of technique and integrity over politics. The steps needed to bring about affirmative action must be taken by the generally autonomous personnel/civil service bureaus. Resistance to affirmative action based on adherence to professional standards would be likely. Affirmative action embodies a political standard to be used in employee selection, a violation of the traditional merit principle which mandated that hiring be based on ability to meet certain job require— ments. Yet, the professional definition of the merit principle had been frequently used as a mask to cover discriminatory practice. The ”scien- tifically-determined” standards often contained "very restrictive minimum qualifications of age, physical characteristics, employment background and training."62 These restrictive qualifications served to prevent many minority applicants from obtaining positions with local government. Nigro and Nigro mention that the qualifications often could not be 18 demonstrably related to ability to perform the job.63 Education requirements may have been arbitrarily high. The years of experience required for certain positions may have been unnecessary given the nature of the work to be performed. In addition, standardized tests may have been required which asked questions not directly pertinent to job performance. Yet, the qualifications had often been in use for many years, were in congruence with a well-developed classification scheme, and provided easy means of numerical computation and assessment. The fact that existing standards barred a disproportionate number of minority applicants was not considered a problem. ”Under traditional civil service, the very low representation of minority groups in the public service was generally not viewed as a violation of the merit prin- ciple.”64 In fact, Nigro and Nigro feel that ”many public personnel workers who are steeped in civil service tradition will continue to resent specific programs for minority groups, women, veterans, and the .65 underprivileged. No one would question that the basic function of a civil service or personnel department is to recruit competent employees for public service, employees selected on the basis of qualifications, of merit. However, specification of appropriate qualifications and a definition of merit are subjective and, hence, subject to differing interpretations. When goals, such as employment by merit, are ambiguous, Scott says that, we see these situations as occasions for the creation of institutionalized organizations . . . . Ritually defined categories can provide order and meaning and rational myths can supply rationales for choice and action. Widely understood roles . . . can offer a basis for legitimate action in the absence of specific organiza- tional guidelines and controls . . . . More specific, proximal, and often procedural goal statements . 66' provide a basis for making decisions and achieving order. 19 The scientific techniques developed by public personnel professionals, their ritualized categories and methods, have become the basis for decision-making in many local personnel offices. The rules, norms and procedures can become a raison d'etre for the bureau, a defense against outside interference and a barrier to change, here to adaptations required to increase minority employment. The value system of many local public personnel administrators has been shaped by long-standing practices, deemed scientific and meritor- ious. New standards of behavior, not congruent with existing practice, will be resisted. As March and Olsen have noted, An organization participant likes what he likes. Indi- viduals come to any particular choice situation with a set of values, attitudes and opinions. These values are substantially fixed. Changes that occur within a relatively brief time period must attend to problems of consistency with the pre-existing attitude struc- ture. In some cases, the restrictions Emposed by this presumption will dominate the behavior. Certainly, affirmative action requirements posit change in conflict with existing value systems. Further, the more massive the change, the greater the reluctance to change is likely to be. Downs said that "it is easier to adjust actions than rules, easier to shift rules than change structures, and easier to alter structures than adopt new pur- poses."68 To make the changes required by federal affirmative action orders, local personnel bureaus mu§£_realign their purposes and funda- mentally adjust their conception of the merit principle. They must also alter methods, procedures, rules and actions, e.g., recruitment strategy, standardized testing, classifications and minimum qualifications, and methods of selection. 20 Therefore, change by these professional bureaus in the face of affirmative action mandates is not likely to be easy. Further, those bureaus with greater autonomy will be the least amenable to adaptation and possess a structure more impervious to political direction, local or federal. Minority Chief Executives Political leaders have often tailored their decision-making to benefit their principal constituencies as rewards for support or to ensure support in future electoral contests. Their decisions may also be tempered by their perceptions of appropriate governmental action, for example, whether there is a need to address certain social inequities or to provide for the enrichment of life in some fashion. Both of these conceptions will affect the actions of municipal executives when they consider adoption of various programs or changes to existing policy and procedures. Because affirmative action is a controversial policy with strong opponents of the majority group, elected or otherwise, it is likely that receptivity to the increased hiring of minority group members will be greater in cities where the elected executive or the appointed executive is a minority group member. Such receptivity is not unique to minority group office holders nor is it necessarily brought about by an affirma- tive action mandate. Rather it is part and parcel of a lengthy tradition in urban America. 21 In many cities, various ethnic groups, most notably the Irish, were able to penetrate municipal workforces as their representatives achieved local elected office. A number of writers in the pluralist tradition have described this ethnic politics phenomenon which is also accompanied by increased personal attention by city hall officials to ethnic group members. For example, in his study of New Haven, Robert A. Dahl noted that between 1899 and 1960 every Democrat successful in his quest for the mayoral office had been an Irish Catholic.69 Concomitantly, the representation of the Irish among city jobholders had risen until "by 1933, the Irish had become by far the most numerous in holding city jObBen70 Wolfinger reported similar findings in his work on New Haven.71 Political office and the related representation in the municipal workforce were paths to upward mobility for previously disad- vantaged groups. Lowi's work on New York72 and Wirt's study of San Francisco73 show similar mobility patterns for the Irish, Italians, and in New York, the Jews, in this century. Although the ethnic politics phenomenon began in an age of patronage rather than merit, it did not-as Lowi and Wirt show-end in that period. In fact, very recent studies have shown that black hiring for city jobs increased during Carl Stokes' brief mayorality in Cleveland and during the first years of Richard Hatcher's tenure in Gary.74 Minority mayors may view increased minority hiring as a way to provide a better quality of service to minority residents and to lessen exacerbated relations between minorities and representatives of authority in addition to strengthening their political base. This is similar to 22 the increased attention enjoyed by the Irish and other groups when elected office also resulted in increased job holding. (Of course, it is possible as well that white mayors may perceive greater hiring of minority workers in the same context.) Further, it should be noted that the presence of a minority mayor may provide an unconscious as well as perhaps conscious spur to the hiring of minorities. Saltzstein has reported on the positive effect of a female mayor on the hiring of women for municipal jobs even if the mayor has no direct control over the affirmative action apparatus.75 On the local level, three sets of determinants have been laid out which could affect compliance with federal affirmative action orders: political form, the degree of professional personnel bureau autonomy, ethnicity. These factors can serve to differentiate among local governments to discern if there are patterns in the degree to which a representative workforce has been approached. These factors can also be incorporated into a formal model of minority representation in local governmental workforces. Further, these factors may prove to be focal points of analysis as well in the examination of other federal programs or administrative orders, especially those which address social needs. For example, discretionary grants may be pursued by those cities having a structure responsive to socioeconomic cleavages. Also, whether or not a local agency is accountable to political actors (local or federal) could deter- mine the manner in which it institutes programs falling within its 23 purview. The programs could be shaped by existing values and practices, values and practices which may contradict program intent. Finally, the presence of a minority executive could facilitate discretionary program adoption or assist in determining who and how many are served. If the factors discussed here do have explanatory power in relation to the hiring of minorities, they may well be key determinants to success or failure elsewhere as well. 5. 6. 7. 8. 9. 10. ll. 12. l3. 14. 15. Chapter I Footnotes Robert T. Nakamura and Frank Smallwood, The Politics of Policy Implementation (New York: St. Martin's Press, 1980), p. 8. Jeffrey L. Pressman and Aaron B. Wildavsky, Implementation (Berkeley: University of California Press, 1974), p. 143. Nakamura and Smallwood, The Politics of Policy Implementation, p. 14. Ibid.’ p. 15. Paul Berman and Milbrey Wallin McLaughlin, Federal Programs Supporting Educational Change, 8 vols. (Santa Monica: The Rand Corporation, 1974-1978), vol. 4: The Findings in Review (1975): 10-11. Ibido, pp. 14-150 Ibide, pp. 17-200 Ibid., pp. 20-21. See Eugene Bardach, ”On Designing Implementable Programs,” in Current Issues in Public Administration, 2nd ed., ed. Frederick S. Lane (New York: St. Martin's Press, 1982); Nakamura and Smallwood, The Politics of Policy Implementation; Daniel A. Mazmanian and Paul A. Sabatier, Implementation and Public Policy (Glenview, IL: Scott, Foresman and Company, 1983). Jeffrey L. Pressman, "Federal Programs and Political Development in Cities” in Improving the Quality of Urban Management, ed. Willis D. Hawley and David Rogers (Beverly Hills: Sage Publications, 1974), p. 598. For specific examples, see Martha Derthick, New Towns In-Town (Washington, D.C.: The Urban Institute, 1972); Pressman and Wildavsky, Implementation; Eugene Bardach, The Implementation Game (Cambridge: The MIT Press, 1978); Bardach, ”On-Designing (Implementable Programs.” Nakamura and Smallwood, The Politics of Policy Implementation, p. 61. Mazmanian and Sabatier, Implementation and Public Policy, p. 37. Nakamura and Smallwood, The Politics of Policy Implementation, pp. 64-65. Bardach, ”On Designing Implementable Programs," p. 406. 24 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 25 Nijole V. Benokraitis and Joe R. Feagin, Affirmative Action and Equal Opportunity: Action, Inaction, Reaction (Boulder, CO: Westview Press, 1978), p. 1. U.S. Equal Employment Opportunity Commission, Affirmative Action and Equal Employment: A Guidebook for Employers, 2 vols. (Washington, D.C.: 1974), 1:4-5. Ibid., p. 350. U.S. Commission on Civil Rights, For All the People. . . By All the People: A Report on Equal Opportunity in State and Local Government (Washington, D.C.: 1969). Harrell R. Rodgers, Jr. and Charles S. Bullock III, Law and Social Change: Civil Rights Laws and Their Consequences (New York: McGraw-Hill Book Company, 1972), p. 123. Bardach, ”On Designing Implementable Programs,” p. 413. Ibid. On the latter point, see Bardach, “On Designing Implementable Programs,” p. 413. Mary Lenn Miller, ”Implementation of Federal Policy at the Local Level: A Study of Affirmative Action in Municipal Governments” (Paper presented at the 1978 Annual Meeting of the Southwestern Political Science Association, Houston, TX, 12-15 April 1978), p. 6. Ibid., p. 8. Ibid. Ibid., p. 18. Herbert Hill, "The Equal Employment Opportunity Acts of 1964 and 1972: A Critical Analysis of the Legislative History and Administration of the Law,” Industrial Relations Law Journal 2 (Spring 1977). Harrell R. Rodgers, Jr., ”Fair Employment Laws for Minorities: An Evaluation of Federal Implementation” in Implementation of Civil Rights Policy, ed. Charles S. Bullock III and Charles M. Lamb (Monterey, CA: Brooks/Cole Publishing Company, 1984), p. 96. Ibid. Ibid., p. 99. Ibid., p. 111. Mazmanian and Sabatier, Implementation and Public Policy, p. 28. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 26 Pressman, "Federal Programs and Political Development in Cities,” p. 590. Ibid., p. 591. Charles R. Adrian and Charles Press, Governing Urban America, 5th ed. (New York: McGraw-Hill Book Company, 1977), pp. 150-151. Donald K. Stone, Douglas Price and Kathryn Stone, ”City Manager Government,” in Classics of Urban Politics and Administration, ed. William J. Murin (Oak Park, IL: Moore Publishing Co., Inc., 1982), p. 94. On this point, see Robert L. Lineberry and Ira Sharkansky, Urban Politics and Ppblic Policy, 3rd ed. (New York: Harper & Row Publishers, 1978); Clarence N. Stone, Robert K. Whelan and William J. Murin, Urban Policy and Politics in a Bureaucratic Age (Englewood Cliffs, NJ: Prentice-Hall, Inc., 1979); Robert L. Lineberry and Edmund P. Fowler, ”Reformism and Public Policy in American Cities,” in Classics of Urban Politics and Administration, ed. William J. Murin (Oak Park, IL: Moore Publishing Company, Inc., 1982). Edward C. Banfield and James O. Wilson, City Politics (New York: Vintage, 1963), pp. 170-171. Ibid. Charles H. Levine, Irene S. Rubin and George G. Wolohojian, ”Resource Scarcity and the Reform Model: The Management of Retrenchment in Cincinnati and Oakland,” Public Administration Review 41 (November/December 1981): 627. Stone, Whelan and Murin, Urban Policy and Politics, p. 80. Levine, Rubin and Wolohojian, "Resource Scarcity and the Reform Model,” p. 620. Jeffrey L. Pressman, Federal Programs and City Politics: The Dynamics of the Aid Process in Oakland (Berkeley: University of California Press, 1975), pp. 28-30. Lineberry and Sharkansky, Urban Politics and Public Policy, p. 220. Ibid. Adrian and Press, Governing Urban America, p. 159. Ibid., pp. 160-161. Levine, Rubin and Wolohojian, "Resource Scarcity and the Reform Model," p. 620. Lineberry and Fowler, ”Reformism and Public Policy,” p. 125. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 27 Ibid. Ibid., pp. 125-126. Ibid. Levine, Rubin and Wolohojian, "Resource Scarcity and the Reform MOdel,” p0 6270 Theodore J. Lowi, At the Pleasure of the Mayor (New York: The Free Press of Glencoe, 1964), pp. 59-61. Wallace Sayre and Herbert Kaufman, ”Administrators of Line Agencies' in State and Urban Politics, ed. Richard I. Hofferbert and Ira Sharkansky (Boston: Little, Brown and Company, 1971), p. 275. Frederick C. Mosher, Democracy and the Public Service, 2nd ed. (New York: Oxford University Press, 1982), p. 73. Ibid., p. 76. Ibid. Ibid., p. 79. Ibid., p. 80. Felix A. Nigro and Lloyd G. Nigro, The New Public Personnel Administration (Itasca, IL: F.E. Peacock Publishers, Inc., 1976), p. 147. Ibid. Ibid., p. 17. Ibid., p. 316. W. Richard Scott, Organizations (Englewood Cliffs, NJ: Prentice-Hall, Inc., 1981), p. 274. James G. March and Johan P. Olsen, "Organizational Learning and the Ambiguity of the Past,” in Ambiguity and Choice in Organizations, ed. James G. March and Johan P. Olsen (Bergen: Universietsforlaget, 1976), p. 62. Anthony Downs, Inside Bureaucracy (Boston: Little, Brown and Company, 1967), p. 173. Robert A. Dahl, Who Governs? (New Haven: Yale University Press, 1961), p. 38. Ibid., p. 41. 71. 72. 73. 74. 75. 28 Raymond E. Wolfinger, The Politics of Progress (Englewood Cliffs, NJ: Prentice-Hall, Inc., 1974). Lowi, At the Pleasure of the Mayor. Frederick M. Wirt, Power in the City: Decision Making in San Francisco (Berkeley: University of California Press, 1974). Susan Welch and Albert K. Karnig, ”The Impact of Black Elected Officials on Urban Expenditures and Intergovernmental Revenue" (Paper presented at the 1978 Annual Meeting of the American Political Science Association, New York, NY 31 August-3 September 1978), p. 3. Grace Hall Saltzstein, ”External Pressures, Organizational Constraints, and Change: The Case of Municipal Affirmative Action" (Paper presented at the 1983 Annual Meeting of the American Political Science Association, Chicago, IL, 1-4 September 1983). Chapter 2 Specifying a Model of Minority Representation in Municipal Workforces Earlier Models of Minority Representation Though a number of writers have examined the hiring of blacks by municipal governments, initial offerings by and large were confined to examinations of aggregate data with descriptive statistics. In 1982, Peter K. Eisinger published two articles on the issue.1 In each, he used econometric models to explain differences in black representation in city workforces. In his first article, his model purported to test the impact of a black mayor on black city employment. He grounded this idea in the existing literature on ethnic politics, particularly the notion that election of members of a particular ethnic group would bring symbolic and perhaps substantive benefits to the group as a whole. He notes that ”few students of ethnic politics have asked the question: Did ethnics whose representatives achieved office in a particular jurisdiction get anything as a result of that co-ethnics in other jurisdictions who failed to attain representation did not?"2 A comparative examination of black municipal workforce representa- tion in order to probe the extent of employment benefits derivative from having a black chief executive is a significant extension of the ethnic 29 30 politics literature. Eisinger uses an Ordinary Least Squares model to undertake his comparison. His main model is as follows: I Blacks in = City + Residency + Public Sector City Population Requirement Growth Workforce + City Population + Z Black College Growth Graduates in City + Black + Black Voting + Black Population Mayor He uses a stepwise regression on a sample of 43 U.S. cities which are at least 10% black and have populations over 50,000. He also reports results from two other models in which he used black administrators and officials and black city civil service professionals as respective dependent variables. Eisinger did not include variables representative of either governmental structure or personnel bureau autonomy. He believes that, Although today the hiring of municipal employees is governed almost entirely by civil service rules, affirm- ative action doctrines provide a means by which some degree of racial advantage may be insroduced into the most rationalized personnel systems. He does not doubt that a degree of advantage will be introduced. He also appears to assume that a black mayor will be able to have an impact in any city which elects one. He offers Detroit and Atlanta as examples. In each case, a newly elected black mayor was able to replace the personnel administrator and several department heads with black officials. In both cities, hiring of blacks rose. Assuming that all black mayors will enjoy such good fortune may be overly optimistic. In some cities, it might not be possible to replace the personnel directors as quickly or as smoothly. 31 Eisinger may have erred significantly by including the percentage of blaCks in the population as an independent variable. He justifies the inclusion because he feels that a large proportion of blacks in a city should bring about greater black municipal employment independent of the race of the mayor. However, this explanation offers tautological wisdom and may obscure the need to discern further political and administrative causal agents. A better method for incorporating black population would be to use population representation to weight the percentage of black representation in the workforce. Or, representation in the workforce could be subtracted from population representation. Under either of these methods, the researcher would be able to assess minority represen- tation in terms of the stated federal goal, a workforce representative of the population. The latter method allows for a greater degree of varia- tion in the dependent variable. Eisinger's overall model of the total workforce has an R2 of .82 but the contribution of black population to the R2 is .585. An adjusted R2 is not reported. When Eisinger does substitute percent of change in the workforce from 1972-1978 as the dependent variable, his R2 drops to .11 and the contribution of black population to the R2 falls to .01. He attributes the changed results to changes in black population which were not controlled4 instead of exploring deficiencies in his original conception of the dependent variable. Eisinger's work lays a foundation for further modeling because of the importance he gives to a black mayor as an explanatory factor. This idea is well-grounded in the literature and lends itself well to a comparative test. In the model Eisinger uses, the variable appears to 32 have a significant effect. However, he has omitted organizational and political factors and his test is marred by the way he handles black population. Eisinger's second article5 features a regression model that in some ways is reminiscent of the basic model discussed above. He sets out to establish that the state of the local economy will affect the percent- age of blacks on the municipal payroll. If the economy is severely strained, black representation will not grow and may in fact decline because of layoffs which affect those with least seniority (dispropor- tionately black) first. Eisinger also adds two other notions. If the private sector is recruiting blacks, that will have a negative impact on black municipal workforce recruitment. Second, citing Frances Fox Piven, he wishes to test her idea that the presence of unions and white ethnics will stifle the hiring of blacks as these two groups have been opposed to equal employment goals. Regrettably, he does not differentiate among the many types of unions representing city workers. Nor does he differentiate among ethnic groups. In the latter instance, he is limited by the type of data collected by the U.S. Census Bureau which keeps records only of those born in a foreign country, not of the children or grandchildren of European immigrants. His resulting model contains a large number of variables: city size, percentage black, black mayor, percent black college graduates, sunbelt location, unemployment in the city, per capita income change (1970-77), growth in manufacturing jobs (1972-77), public employees per 1000, per capita local government noneducation expenditures, change in 33 percentage of blacks in private sector managerial jobs, public employee union coverage, percent foreign stock. Here again his endogenous variable is percentage of blacks in the city workforce. Using the same sample of cities as his previous work, he estimated his model with Ordinary Least Squares though the stepwise method is not employed here. Eisinger has again used percentage of blacks in a city's population as an exogenous factor. His model does produce an adjusted R2 of .71 but only three of his thirteen coefficients--city size, black population, and unemployment in city-are statistically significant at the .05 level. Several of the coefficients are negative, opposite to what had been hypothesized. In both of his works, Eisinger has drawn attention to certain exogenous, chiefly economic factors, which must be included in any model of black, or minority, municipal employment. His findings on the black mayor are also important, although this variable becomes insignificant in his second work. Eisinger's attempts are the first in the area and certainly provide information to students of affirmative action policy. They are a useful beginning but point out the need for a tighter concep- tualization of how affirmative action occurs in diverse local governments and the need to construct an endogenous variable which can provide for comparative assessment of whether cities are meeting prescribed goals. Initial Dimensions of Analysis Eisinger's work has highlighted certain factors important to an explanation of the degree of minority representation in municipal 34 workforces. Although he does attempt to address the concept of ethnic politics, he does not include variables pertaining to either political form or organizational autonomy. The principal aim of this work is to create an econometric model which will add to the understanding of minority employment in local government, as a whole and within one particular sector, the protective services (police and fire). The key dimensions to be addressed in this effort are those discussed in Chapter 1: political form (the mayor- council vs. the council-manager system), civil service autonomy, the influence of minority chief executives (elected or appointed). Of course, other relevent exogenous factors dealing with the economy and enforcement also will be added. There is a natural dichotomy between the council-manager form of government and the mayor-council form. The former represents administra- tion by an unelected expert, often hired from outside the community and usually interested in managerial capacity and technical expertise rather than social issues.6 Over 95% of city managers are white and are often of middle class origins. Further, the form itself tends to blur distinc- tions based on socioeconomic and racial cleavages. In addition, elected council members serve only in a parttime capacity and often have few resources or staff for the monitoring of administration. In mayor-council cities, the mayor is elected by a coalition of the city's citizens and is generally the head of the administration, with a staff and resource capability to monitor bureaucratic output. The powers of the mayor will vary from city to city depending on the presence of independent commissions. Nonetheless, an elected mayor, a political 35 leader, may be more sensitive to the natural clientele of an affirmative action program. He or she is a local resident, often a native, and may have working class origins. Of course, the composition of the city may affect attention paid to potential recipients of the benefits of affirma- tive action. If the minority vote is critical to electoral success, it is likely that mayors may be especially sensitive to the issue. It can thus be hypothesized that, all other things being equal, the mandate for affirmative action will be received more favorably in mayor- council cities than in council-manager cities. (Refer to Table 2-1). Table 2-1 The Political Form Dimension: Receptivity to Affirmative Action Mayor-Council Cities + Council-Manager Cities Further, mayor-council cities can be differentiated on two grounds: whether council members are elected at large or from wards and whether the contest for mayor and council is a partisan one. It is usually assumed that a city with a ward system and partisan elections would show the greatest degree of response to a program geared to minority groups. Similarly, personnel departments in cities may be grouped in two categories by the degree of autonomy they enjoy. On the one hand, 36 independent civil service commissions should have the greatest freedom to pursue professional practice as their director is often named by a commission whose appointees may not be reflective of the current mayor or city manager. Though these latter officials may have some say in selec- tion, their choice may not always be accepted. Also, many directors and/or commissioners could have career (or lifetime) appointments. On the other hand, a personnel department without a governing commission, whose head is selected directly by the city's chief administrator, is more likely to be subject to executive politically-based direction. It is therefore hypothesized that the greater degree of autonomy and professional practice a public personnel bureau evinces, the less likely it is to accommodate its mode of operation to new political considera- tions, namely the affirmative action order. A civil service commission will have greater autonomy than a personnel bureau directly responsible to the city's executive. (Table 2-2) Table 2-2 The Bureaucratic Autonomy Dimension: Receptivity to Affirmative Action Civil Service Commission Personnel Department 4. 37 A similar division can be made regarding the presence or absence of a minority chief executive. It is hypothesized that the presence of a minority elected mayor will provide a conscious or unconscious spur to minority hiring. In addition, the presence of a minority city manager should bring about a similar effect. Finally, in a number of council-manager cities, there is an independently elected mayor, albeit with perhaps few concrete administrative duties. The-presence of such a minority mayor could create some greater awareness of minority needs and assist in increasing receptivity to affirmative action. The expected results are illustrated in Table 2-3. Table 2-3 The Ethnic Politics Dimension: Receptivity to Affirmative Action Presence of a Minority Mayor and/or City Manager 4. Nonminority Mayor or City Manager These three key local determinants have been arrayed together (Table 2-4) to form eight possible local combinations which can affect the implementation of affirmative action. The combinations represent 38 hypothesized interactions but do not estimate the strength of each factor. However, they do provide an initial means for grouping cities in order to see whether these sets of localized circumstances provide an appropriate means for distinguishing degrees of minority representation. They are not absolutes because there are other exogenous factors which can affect receptivity to affirmative action and minority hiring. Further, they do not permit gradations based on an elected mayor's political base, the opportunity or ability of a mayor or city manager to appoint a new personnel director or the legitimacy afforded affirmative action by key actors. Nonetheless, the typology is a starting point for analysis of differences among cities prior to the creation of a formal model. Table 2-4 Local Determinants of Affirmative Action Implementation Civil Service No Civil Service Commission Nonminority ___-_-_-________________-;+:____ COUNCIL- MANAGER Minority --+ —++ Nonminority __:-_-_-_______________:+:_____ MAYOR- COUNCIL Minority ‘ +—+ +++ Key: Political Bureau Minority Form Autonomy Executive 39 Initial Methods of Analysis The Sample Cities included in this study are those with populations of 100,000 and over and which are at least ten percent minority. All minorities considered protected classes under U.S. law are included: blacks, Hispanics, American Indians, Asians, Alaskan natives and Pacific Islanders. All these groups have been underrepresented to a greater or lesser extent in municipal workforces.7 Hence, the adoption of changes to facilitate affirmative action should benefit them all. Further, the traditional public personnel practices which disproportionately kept blacks outside municipal workforces had a negative effect on other non- majority groups as well. One hundred forty-five mayor-council and council-manager cities were surveyed by mail in November 1982. They were asked whether or not the city had a civil service commission. They were also asked to report the following information regarding their 1982 workforce composition (taken from the annual report filed by each city with the U.S. Equal Employment Opportunity Commission): 1. Total number of employees 2. Total number of white employees 3. Total number of protective service (police and fire) employees 4. Total number of white protective service employees Data has been collected separately on protective service employees because, historically, police and fire departments have been judged the most resistant to the hiring of minorities.8 In 1969, the U.S. 40 Commission on Civil Rights found that although 27% of all city employees were police officers and firefighters, only 7Z of black employees were.9 Underrepresentation was acute in uniformed jobs and very few minorities were found above the rank of patrolman in police depart- ments.10 Underrepresentation was found to be even greater in fire 11 departments. Sample summary findings are contained in Table 2-5. Table 2-5 Minority Representation in Central City Protective Services, 1967 San Francisco Oakland Philadelphia Detroit Z Black--all city functions 18.5 15.3 40.6 40.1 Z Black-police protection 4.9 5.2 24.0 10.1 Z Black-~fire protection .1 4.3 7.8 3.8 Z Asian-all city functions 4.3 1.6 -— -- Z Asian-~police protection .2 .2 -- - Z Asian-fire protection .1 0.0 - - Source: U.S. Commission on Civil Rights, For All the People, pp. 11 and 28. Judging from earlier studies, particularly that of the U.S. Commis- sion on Civil Rights, it appears plausible that overall gains in minority participation in municipal employment may not be evenly distributed across all city departments. As police and fire were documented to fare 41 most poorly in earlier years, it was decided to maintain individual statistics on the protective services so that their employment patterns might be analyzed separately along the dimensions outlined above. If distinct differences were discovered, it would also be possible to study the workforce minus its protective services component. Figures on total minority population were taken from the 1980 U.S. Census. As the acknowledged goal of affirmative action is a workforce representative of a city's population, a ratio has been constructed to measure achievement in terms of that goal, namely Z Minority in Work- force/Z Minority in Population. In addition, the percentage difference between minority representation in workforce and in population has been calculated. These provide methods of comparison across cities by controlling for the logical notion that if a city has more minorities, it will have more minority workers and allows for better analysis of political and organizational factors. One hundred thirty-four cities of the possible one hundred forty- five responded to the survey. A list of those cities, showing minority population and minority workforce representation, is located in the Appendix. Two cities indicated that they were unable to respond because of their current involvement in anti-discrimination suits and a third declined to provide information of such a sensitive nature. Analysis of Variance The data set will be broken down by political form, autonomy of the personnel bureau, and the presence of a minority chief executive. Each 42 dimension will be arrayed separately as well as in conjunction, such as in Table 4. Further, mayor-council cities will be grouped on the basis of whether or not they have partisan mayoral elections and whether elec- tions for council are at large, by wards or a combination of the two methods. The variable to be compared and analyzed in each instance will be the percentage gap between minorities in the workforce and minorities in the population. In addition, analysis will be done using the gap between percentage of minorities in the protective services and percentage of minorities in the population. Difference of means tests will be conducted among the cells of the various groupings in order to determine the significance of the three key local determinants highlighted here. Further, initial judgments can be made regarding minority representation in protective services, an area of key interest to minority city dwellers. Does representation in this area follow the same pattern as in the total workforce? Do the key local factors work in the same manner for this particular segment of the work- force? It is hoped that significant differences can be found among cities because of political form, personnel bureau autonomy and ethnic politics, pursuant to the literature. These tests are a useful preliminary means of evaluating one's assumptions; of course, other exogenous factors are not entered into the calculations at this point. Therefore, also prelim- inary to model construction, Pearson Correlations and partial correla- tions will be calculated in order to gain preliminary indications of the effects of economic conditions on city responsiveness to affirmative action. These results will also be discussed prior to model estimation. 43 The Model An econometric model will be constructed, improving on earlier efforts by Eisinger, and grounded in existing research on organizations and political form. Political and administrative factors will be utilized in order to pinpoint principal determinants of local implementa- tion of a federal policy. It is the contention of this research that a city's success in increasing the minority presence in its workforce is a product of the degree to which it can be held accountable to federal political affirma- tive action orders. This will depend both on how autonomous its person- nel bureau is and how amenable its form of government may be to implemen— tation of such directives. In addition, the presence of a minority elected mayor and/or minority city manager may increase the pace of minority hiring, either consciously or unconsciously. In order to fully specify the causal factors at work, local economic conditions must be included. Constrained municipal budgets, products of the recent severe recession, may have led to sharp hiring curtailment or to layoffs. In the latter case, adherence to seniority rights may have caused the dismissal of a disproportionate number of minorities, those hired most recently. Three variables have been created to tap into this economic component. The first is the average unemployment rate in each city for the year 1982, the year of highest unemployment in the U.S. since the 19308. High unemployment should have a negative effect on minority hiring. Second, the growth or decline in workforces from 1977 to 1982 will be included. (Until 1977, workforces had been growing 44 rather steadily.) This will be operationalized as a ratio, number of employees in 1982 divided by number in 1977. As some temporary Compre- hensive Employment and Training Act employees may have been included in the 1977 figures (taken from the U.S. Census of Governments) and are not included in the 1982 survey responses, those cities that actually did increase their workforces should show a greater number of minority employees, all other things being equal. Finally, in the economic area, much has been made of the overall division between Sunbelt and Frostbelt, between growth and stagnation. To control further for the strained circumstances of municipal govern- ments in the Frostbelt, a dummy variable will be entered into the model, reading one for a city outside New England, the Northeast and the Mid- west. This variable will also, to some extent, be a surrogate for popu- lation migration to warmer climates which the 1980 census demarcated. Although federal enforcement efforts in the area of affirmative action have not been constant or certain, it is nonetheless likely that federal threats to withhold funds or court suits related to equal employ- ment may serve as a prod to cities to adjust their recruitment and selec- tion patterns. Data regarding threats to withhold funding was unobtain- able from certain agencies of the federal government, i.e., Department of Housing and Urban Development, Office of Management and Budget, Equal Employment Opportunity Commission. Reasons given included the fact that such data were stored at many different sites and would be difficult and expensive to compile. However, the Civil Rights Division of the Justice Department did stipulate that no city had had funds terminated because of failure to comply with affirmative action orders. The Civil Rights 45 Division was able to provide a listing of all cities in the sample which have been involved in employment discrimination suits brought by the Justice Department. In addition, the Legal Defense Fund, a nonprofit civil rights organization, provided information regarding suits it or other private parties had initiated. Although Eisinger found its effect was negligible,12 whether or not a city has a residency requirement for muncipal workers has been included as an independent variable. If a residency requirement is present, minorities concentrated in central cities should have an advantage. The following model will be estimated, using Ordinary Least Squares: Z Minority in City = Political Form + Bureaucratic Autonomy Population - Z Minority (0,1) (0,1) in City Workforce + Minority Executive + Unemployment Rate (0.1) + Workforce Change + Sunbelt (0.1) + Enforcement + Residency Requirement (0.1) (0,1) The endogenous variable again is the percentage gap between popula- tion and workforce representation. This moves minority population to the left side of the equation so that cities may be assessed by the degree to which they attained a stipulated goal. This model will also be regressed on the percentage gap between minority population and minority representation in the protective 46 services. If the protective services model offers very different results from the workforce as a whole a third model will be added covering the overall workforce minus the protective services component. It may be that police and fire, normally professionalized and highly autonomous bureaus, would prove more resistant to recruitment and selection changes. Because of their relative insularity, these bureaus may be more affected by court suits than by political and organizational factors. In that case, separate models would be justified to capture the disparate forces at work. It should be noted here that the protective services are frequently cited in the literature (notably in the writings of Sayre and Kaufman and of Lowi) as examples of autonomous, professional bureaus resistant to executive direction.13 Analysis of Results The models developed will be assessed in terms of whether the assumptions of the general linear model have been fulfilled. Diagnostic tests are available and correction techniques exist. The models will be judged by whether their coefficients are statis- tically significant and in the direction posited and goodness of fit measures will be examined. Results will also be compared to those of Eisinger, though of course not all variables will be the same. It is hoped that the models created will shed significant light on key determinants in local government which can influence the implementation of affirmative action and other federal policies and programs. The models will not purport to explain the implementation 47 process in every city nor in any particular city but should highlight key common factors which will be useful for further examinations of this type. 2. 3. 8. 9. 10. ll. 12. 13. Chapter II Footnotes Peter K. Eisinger, ”Black Employment in Municipal Jobs: The Impact of Black Political Power,” American Political Science Review 76 (June 1982); Peter K. Eisinger, ”The Economic Conditions of Black Employment in Municipal Bureaucracies,” American Journal of Political Science 26 (November 1982). Eisinger, ”Black Employment in Municipal Jobs," p. 381. Ibid. Ibid., p. 389. Eisinger, ”The Economic Conditions of Black Employment." Lineberry and Sharkansky, Urban Politics and Public Policy, p. 220. U.S. Commission on Civil Rights, For All the People. Ibid., p. 16. Ibid. Ibid. Ibid., p. 17. Eisinger, "Black Employment in Municipal Jobs,” p. 387. Sayre and Kaufman, ”Administrators of Line Agencies” and Lowi, fit_ the Pleasure of the Mayor. 48 Chapter 3 Initial Statistical Findings Concerning Local Political and Organizational Factors It is not the contention of this research that political and organi- zational factors are the sole or even the principal explanation for vari- ation in achievement of a workforce representative of population. Economic and demographic factors have been demonstrated in Eisinger's previous work to have considerable power. Nonetheless, political and organizational inputs are not to be dismissed nor discounted. In order to establish the salience of these factors, their effects on workforce representation are explored first, individually and in combination. Then, controls for economics and demographics will be applied in correla- tional analysis to see if the effects of political form, bureaucratic autonomy or presence of a minority executive survive the influence of established determinants from other spheres. Before beginning analysis along political and organizational dimen- sions, it is useful to examine how cities fared as a whole in achieving a representative workforce. The statistic used to judge progress toward goal attainment is the percentage of a city's minority population minus the percentage of minorities in the city's workforce. A positive number indicates a gap between the percentage of minorities in the population and minorities in the workforce, i.e., a shortfall in terms of attainment of the federally-prescribed goals. On the other hand, a negative number represents a surpassing of the goals. The results can be found in Table 3-1. 49 50 Table 3-1 Average Shortfall in the Achievement of a Representative Workforce STANDARD N OF MEAN DEVIATION CASES Total Workforce 2.18Z 10.31Z 134 Protective Services 13.11Z 10.93Z 134 Workforce Without Protective Services -2.63Z 12.28Z 134 Contrary to what has been expected, it does appear that many cities have made substantial progress in achieving a workforce representative of municipal population. An average shortfall of only slightly over two percent indicates general compliance though of course an examination of the raw data for the sample cities (Appendix) indicates a good bit of variation along certain lines which can be examined statistically here. The 13.11Z shortfall for the protective services supports the contention of the U.S. Commission on Civil Rights that police and fire departments are underrepresented even when workforces as a whole employ a significant number of minorities. When the protective services sector is removed from the workforce, the remaining portion, on average, has minority representation actually in excess of representation in the city popula- tion. These findings can be illustrated in another manner by using a different statistic: a ratio of percent minority in the city workforce divided by percent minority in the city's population. Here, 1.00 is equivalent to attainment of prescribed representation. A number greater than 1.00 represents a surpassing of the federal goals while a number less than 1.00 indicates underrepresentation. (See Table 3-2.) 51 Table 3-2 Mean Ratio of Minority Representation STANDARD N OF MEAN DEVIATION CASES Total Workforce .949 .327 134 Protective Services .595 .256 134 Workforce Without Protective Services 1.110 .436 134 Again we generally see evidence of considerable progress by cities in the area of minority hiring. However, despite these generally posi— tive results regarding goal attainment, it is still possible to judge whether local political and organizational characteristics contributed to the degree of success or failure. It should also be noted here that in regard to the protective services, fire and police departments began with the worst minority hiring records and thus had a more difficult target to reach. In addition, their status as generally autonomous, professional organizations (sometimes with their own civil service systems), should work against accountability to goals prescribed outside their profes- sion. Analysis of Variance According to Key Local Dimensions Political form The first local determinant discussed in this work was political form. It was hypothesized that council-manager cities, which are regarded as less responsive to socioeconomic and racial cleavages, 52 would be less likely to achieve representative workforces than mayor- council cities. However, as Table 3-3 shows, the results have been just the opposite. The statistic used in this and the subsequent analyses is the short- fall between minority representation in a city's population and minority workforce representation. Again, a positive number indicates a failure to achieve the prescribed goal of a representative workforce while a negative number is indicative of the surpassing of the federal target.- Data arrayed in Table 8 and subsequent tables will also include separate entries for the protective services and the workforce without protective services. These have been included because the mean shortfall for the protective services was considerably larger than for the workforce as a whole. Table 3-3 Minority Workforce Representation by Political form N OF SIGNIFICANCE MEAN CASES Z LEVEL TOTAL Council-Manager .36 76 WORKFORCE (8.90) -2030 005 Mayor-Council 4.56 58 (11.57) PROTECTIVE Council-Manager 12.17 76 SERVICES (8.74) -1.08 .30 Mayor-Council 14.34 58 (13.25) WITHOUT Council-Manager -4.38 76 PROTECTIVE (11.41) Mayor-Council -.34 58 (13.09) 53 For all three categories depicted in Table 3-3, council-manager cities perform better, in terms of goal achievement, than their mayor- council counterparts. These results begin to raise some doubt as to the purported inability of council-manager cities to address racial cleav- ages, at least in terms of employment opportunities. Of course, no control variables have been introduced as yet and it is conceivable that economic circumstances might affect the efforts of the large number of mayor-council cities found in the Northeast and Midwest. Further analysis of variance will occur along the lines of bureau- cratic autonomy and presence of a minority mayor/manager. Later use of correlational analysis will allow for the placement of economic controls. Finally, as council-manager cities are generally less populous and have smaller minority populations than the mayor-council cities in the sample, controls for these factors will be introduced as well. Bureaucratic Autonomy It has been posited here that bureaucracies are reluctant to change and that those bureaus which are both professional and autonomous are able to resist change most successfully. Therefore, the presence of a formal civil service commission was hypothesized to be an impediment to achievement of a representative workforce. Such independent bureaus would be more likely to and could better resist political changes to long established practices. The results of a breakdown of cities by whether or not they have a civil service commission are found in Table 3-4. 54 Table 3-4 Minority Workforce Representation by Bureaucratic Autonomy N OF SIGNIFICANCE MEAN CASES Z LEVEL TOTAL Civil Service 3.50 84 WORKFORCE (8.43) -1.77 .10 None -.05 50 (12.64) PROTECTIVE Civil Service 13.61 87* SERVICES (10.44) -.69 .50 None 12.18 47 (11.85) WITHOUT Civil Service -.97 84 PROTECTIVE (10.35) SERVICES -1.88 .10 None -5.42 50 (10.35) *Several cities have civil service commissions only for their police and fire employees. Here the expected difference between cities which have civil service commissions and those that do not is clearly in the direction hypothe- sized. However, the significance of the difference weakens considerably when the protective services is examined as a separate unit. This may be partially a result of the protective services themselves being autonomous professional units. Their role in job selection could be greater than that of other city departments where a civil service commission is present. 55 The Presence of a Minority Executive Drawing upon the ethnic politics tradition, it was hypothesized that greater minority workforce representation would be achieved in those cities which have minority elected mayors and/or minority city managers. Table 3-5 depicts the results of this bivariate breakdown. Table 3-5 Minority Workforce Representation by Presence of a Minority Executive N OF SIGNIFICANCE MEAN CASES Z LEVEL TOTAL Nonminority 2.21 117 WORKFORCE (9.81) .07 .95 Minority 1.97 17 (13.64) PROTECTIVE Nonminority 12.44 117 SERVICES (10.59) 1.67 .10 Minority 17.72 17 (12.43) WITHOUT Nonminority ~2.52 117 PROTECTIVE (11.66) SERVICES .36 .75 Minority -3.36 17 (16.38) The findings in Table 3~5 again were unanticipated. For both the total workforce and the workforce minus protective services the differences between minority- and nonminority-governed cities are in the direction hypothesized. However, for all intents and purposes the differences found are most likely due to chance. It is true that the 56 difference in the number of cases in the cells of this bivariate break- down could affect whether there are findings of significant differences. But, here this appears unlikely. Protective Services again presents a different pattern. Here, a significant difference between cities with minority and nonminority exec- utives does appear. But, it is the cities which have minority executives which experience the greatest shortfall between minority population and minority employment in police and fire departments. A number of possible explanations can be offered for both sets of unanticipated findings. For one, cities with minority executives are very often those cities experiencing population loss and general economic decline. A number, with Detroit as a prime example, were most hard hit by the recent recession and suffered cutbacks in city employment. Controlling for economic factors may increase the negligible differences found between those cities with and without minority executives in the total workforce and the workforce without protective services. Of course, it must be again noted that a surprising proportion of all sample cities came close to or exceeded achieving a workforce representative of population. (See Appendix.) The economic factors cited above of course will affect the protec- tive services results as well. In addition, it is likely that large cities with minority mayors have well-established police and fire unions with collective bargaining powers. These unions, most often representa- tive of the desires of white officers, may have attempted to stymie efforts to promote integration. After repeated attempts, it was found that there is no available list of police unions with collective 57 bargaining power, even from the American Organization of Police Associations. Effort will continue to assemble such a list but at this time this factor cannot be utilized as a control mechanism. Key Determinants in Pairs In order to further analyze the variation among cities, the three organizational and political determinants have been paired, i.e., political form and bureaucratic autonomy, political form and minority executive, bureaucratic autonomy and minority executive. The results of the first pairing, political form and bureaucratic autonomy, are found in Table 3-6. Table 3-6 Minority Workforce Representation by Political Form and Bureaucratic Autonomy Manager/Council Mayor/Council TOTAL None -2.07 4.25 WORKFORCE (10.05) (16.45) 34 16 Civil Service 2.33 4.68 (7.40) (9.32) 42 42 PROTECTIVE None 11.52 13.58 SERVICES (9.90) (15.52) 32 15 Civil Service 12.64 14.60 (7.87) (12.56) 44 42 WITHOUT None -7.12 -1.82 PROTECTIVE (12.32) (18.58) SERVICES 34 16 Civil Service -2.16 .23 (10.15) (10.53) 42 42 58 For the protective services, the relationships depicted in Table 3-6 are in the direction expected from the earlier arrays. However, there are no statistically significant differences among the cells. For the workforce as a whole, the patterns of the bivariate break- downs are also repeated. However, among the council-manager cities there is a statistically significant difference at the .05 level between cities which have civil service commissions and those which do not. As expected, the cities without civil service have achieved greater minority workforce representation. This significant difference is not found among the mayor-council cities. Significant distinctions can also be found between the mayor-council cities and their council-manager counterparts. Council-manager cities again fare better in achieving greater minority workforce representation. When there is no civil service commission present, the difference in favor of the council-manager cities is significant at the .15 level. When the cities have civil service commissions, the difference is in the same direction and is significant at the .20 level.1 When protective services is deleted from the workforce, the results parallel those of the workforce as a whole except that the levels of significance are lower. Among council-manager cities, the difference between those cities with and those cities without civil service commis- sions is significant at the .10 level, again with higher goal attainment on the part of the non-civil service cities. There are significant differences at the .30 level between mayor-council and council-manager cities, both for those with and without an autonomous civil service. Again, the council-manager cities fare better. 59 The presence or absence of a civil service commission appears to affect council-manager cities to a greater degree than it does mayor-council cities. This conclusion does not, however, apply to the protective services. The next second-level breakdown of workforce representation to be examined involves political form and the presence of a minority elected mayor and/or city manager. The results are found in Table 3-7. Table 3-7 Minority Workforce Representation by Political Form and Presence of a Minority Executive Council- Mayor- Manager Council TOTAL Nonminority .27 4.80 WORKFORCE (7.54) (11.81) 67 50 Minority 1.01 3.05 (16.53) (10.51) 9 8 PROTECTIVE Nonminority 11.81 13.28 SERVICES (8.61) (12.81) 67 50 Minority 14.87 20.93 (9.71) (14.94) 9 8 WITHOUT Nonminority -4.67 .35 PROTECTIVE (9.72) (13.41) SERVICES 67 50 Minority -2.22 -4.64 (20.85) (10.62) 9 8 60 In the second-level breakdown depicted in Table 3-7, there is a statistically significant difference at the .05 level between nonminor- ity-governed mayor-council and council-manager cities, for the workforce as a whole and also for the workforce excluding protective services. Again, council-manager cities fare better in achieving a representative workforce. Although the difference is not statistically significant, it is interesting to note that in these arrays, council-manager cities with nonminority executives do better than those governed by minorities. For the total workforce and the workforce minus protective services, the differences that exist among mayor-council cities are in the hypothe- sized direction. Minority-governed cities show greater progress although here too the differences are not statistically significant. The protective services present a different twist. In this grouping, there is one significant finding: mayor-council cities with minority mayors have a greater shortfall between goals and actual minority employment than cities with nonminority mayors. The finding is significant at the .20 level but again it is possible that economic factors may be responsible for this unexpected result. The final second-level breakdown examines the effects of bureau- cratic autonomy and presence of a minority executive on municipal employ- ment. The results are contained in Table 3-8. In every cell displayed in Table 3-8, those cities which do not have civil service commissions exhibit more extensive minority participation in the workforce. This is in keeping with an important research hypothe- sis. Two of the relationships are statistically significant. For both the total workforce and the workforce without protective services, the 61 Table 3-8 Minority Workforce Representation by Bureaucratic Autonomy and Presence of a Minority Executive No Civil Civil Service Service TOTAL Nonminority .20 3.42 WORKFORCE (12.14) (7.95) 44 73 Minority -1.88 4.07 (17.12) (11.72) 6 11 PROTECTIVE Nonminority 12.03 12.66 SERVICES (11.92) (9.87) 41 76 Minority 13.19 20.19 (12.39) (12.31) 6 11 WITHOUT Nonminority -5.37 -.81 PROTECTIVE (13.89) (9.79) SERVICES 44 73 Minority -5.84 -2.01 (21.29) (14.03) 6 11 difference between nonminority-governed cities with and without civil service is significant at the .12 and .10 levels respectively. In these arrays, minority-governed cities fare better though not at a significant level. However, in examining only the protective services, it is nonminority-governed cities which have achieved more representative workforces. In fact, for cities with civil service commissions, the difference between them is statistically significant at the .10 level. Thus far it appears that council-manager cities have been more successful in achieving workforces representative of city population, 62 contrary to expectations. However, as hypothesized, the presence of formal civil service commissions, independent professional bureaus, does seem to result in fewer minority gains. The only significant finding involving the presence of a minority executive is that of a negative effect on hiring in the protective services. What now remains is to group minority representation in the sample cities by all three local factors. The Complete Typology As a last effort to analyze variance in minority employment in the sample cities, the data were grouped according to the tripartite typology specified in Table 2-4. The results, which are shown here in Table 3-9, continue the basic patterns that have already been highlighted. In this finer breakdown, it becomes more apparent that council- manager cities with minority executives fare less well than their non- minority counterparts. The differences are not statistically signifi- cant, however, and may be attenuated when controls for the economic climate are applied. In the workforce as a whole, two significant relationships are main- tained: between nonminority-governed council-manager cities with and without civil service commissions (at the .01 level), and, between non- minority-governed mayor-council and council-manager cities without civil service commissions (at the .10 level). Not surprisingly in the first case, the non-civil service cities show greater minority workforce 63 Table 3-9 Minority Workforce Representation by Political Form, Bureaucratic Autonomy and Presence of a Minority Executive WITHOUT PROTECTIVE PROTECTIVE TOTAL WORKFORCE SERVICES SERVICES No Civil Civil No Civil Civil No Civil Civil Service Service Service Service Service Service M N (8.24) (6.38) (9.39) (8.11) (10.11) (8.87) A 29 38 27 4O 29 38 G E Minority -062 3.04 14.91 14.81 -3092 -095 R (4.34) (15.67) (13.02) (5.05) (23.22) (20.73) 5 4 5 4 5 4 M White 5.08 4.69 14.22 12.91 -.92 .89 A (16.68) (9.29) (15.90) (11.64) (18.86) (10.57) Y 15 35 14 36 15 35 O R Minority -8.18 4.66 4.58 23.27 -.15 -3.11 (0) (10.24) (0) (14.47) (0) (10.46) 1 7 1 7 l 7 participation. In the second, council-manager cities fare better than their mayor-council brethren. These relationships are maintained when the protective services units are removed from the workforce although the level of significance drops (to .05 and .20 respectively). Within protective services, the curious relationship involving the presence of a minority executive which was noted earlier still pertains. There is a statistically significant difference at the .10 level among mayor-council cities with civil service commissions between those with minority mayors and those without. Again, it is the minority-governed cities which are farther from achieving workforces representative of their population. 64 The various groupings discussed thus far appear to confirm one key research hypothesis, namely the negative effect of bureaucratic autonomy on minority workforce representation. The findings, however, must still be construed as tentative because other exogenous factors have not been taken into account. Following a brief discussion of differences within only mayor-coun- cil cities, other factors relevant to minority workforce representation will be controlled for, first through the use of correlations and then in econometric models. The Mayor-Council Cities: Variations Due to Partisanship and the Nature of Council Elections The growth of council-manager cities was a clear product of the Progressive reform movement. The professional expert was to become the municipal administrator in an attempt to separate politics from admini- stration. However, the reform movement also led to certain changes in many of those cities which retained an elected mayor as chief administra- tive officer. A number of cities instituted nonpartisan elections for mayor to curtail the influence of political parties. Another change adopted by many cities was the decision to hold at-large elections for the city council. This change was a reaction against the ward bosses who dispensed patronage and operated an informal network of social services in exchange for votes. These Progressive era changes are reflected today in the continuing differences in electoral rules among mayor-council cities. Those cities 65 which retain partisan elections for mayor and district council elections are felt to be the most unreformed, the most politically responsive. In the most reformed cities, those with nonpartisan and at-large elections, it has been felt that certain groups were not permitted adequate representation. "Unpopular elements” such as blacks or labor candidates ”were frequently denied formal electoral representation because of nonpartisan, at-large elections” and other reforms.2 ”There was no incentive to try to represent the various ethnic, racial, and religious groups in the city."3 However, in Chicago, a ward city with partisan elections, blacks were able to gain representation on the city council at a much earlier date.4 Given this history, it can be hypothesized that cities with the most unreformed structure would be more likely to take into account minority needs and desires and thus accelerate minority hiring. In order to examine this question, data for mayor-council cities have been grouped by the nature of their electoral rules (Table 3-10). Table 3-10 Effects of Partisanship and Type of Council Elections on Minority Workforce Representation in Mayor-Council Cities WITHOUT PROTECTIVE PROTECTIVE TOTAL WORKFORCE SERVICES SERVICES Council Non- Non- Non- Elections Partisan Partisan Partisan Partisan Partisan Partisan At-large 5.44 3.20 19.03 11.89 .63 -l.88 (14.46) (2.96) (15.39) (2.95) (15.39) (4.95) 10 5 10 5 10 5 At-large 4.87 5.13 18.58 14.21 1.75 -.22 and (13.69) (7.80) (13.39) (11.86) (13.78) (13.25) Districts 9 12 9 12 9 12 Districts 3.35 4.85 10.27 12.08 -.33 .58 (9.55) (12.47) (16.52) (10.99) (11.70) (16.80) 12 10 12 10 12 10 66 The results show very few differences among the mayor-council cities. The most apparent variation can be found in examination of the protective services as an isolated unit. And, here, we see results reflective of the literature cited earlier, at least in respect to the positive influence of partisan elections for mayor on minority hiring. However, due to large variances and a small number of cases in each cell, very few differences which appear are statistically significant. One exception is the difference between partisan and nonpartisan cities with at-large council elections for the protective services component only (.25 level). From these results, it appears that the further divisions undertaken here add very little explanatory power. These findings again contradict some of the standard writings on governmental structure and municipal electoral rules. Correlational Analysis It has been posited above that the relationship of key local deter- minants-~political form, bureaucratic autonomy, the presence of a minority mayor and/or city manager--may be affected by other exogenous factors, principally economic. The recent severe economic downturn, beginning in certain areas such as the industrial midwest in 1979, affected some locales more strongly than others. It is quite likely that in those areas where the recession was most marked, municipal hiring was curtailed and some workers may have been laid off. Hiring cutbacks or 67 workforce diminution would adversely affect efforts to achieve a municipal staff representative of city population. These economic factors will have a direct impact on minority hiring as well as tempering the effects of other variables. The effects of economic factors on key local determinants can first be examined by using Pearson correlations and partial correlations. In Table 3-11, the gap between minority population representation and minority representation in the total workforce has been correlated with political form, bureaucratic autonomy and presence of a minority executive. Controls are also introduced for the average 1982 unemployment rate in the sample cities and for increase or decrease in municipal jobs between 1977 and 1982. Table 3-11 Correlations between Minority Representation in the Total Workforce and Key Local Determinants Z Minority in Population - Z Minority in Workforce Political Form .203 Bureau Autonomy .167 Minority Executive -.008 Control for Unemployment Rate .089 .151 -0075 Control for Workforce Change .164 .161 -.033 Control for both .077 .150 -0 082 Owing to the form of the dependent variable, a positive correlation is indicative of a negative effect on achievement of a representative workforce. In keeping with earlier analysis, political form is 68 positively correlated with the gap between population representation and workforce representation. Political form, a dichotomous variable, is coded 1 for mayor-council cities. However, the strength of this rela- tionship is cut by more than half when economic controls are applied (from .203 to .077). The significance of structural differences is certainly attenuated by the addition of other factors. The advantage enjoyed by council-manager cities may no doubt be dissipated in a model. The negative effects of an autonomous, professional civil service are maintained in this series of correlations. Of special interest is the fact that economic controls only mildly temper the relationship (from .167 to .150). It is more likely that the negative effects of bureau- cratic autonomy will remain present amidst a variety of local circum- stances. To that end, it is interesting to note that the correlation between a sunbelt location and presence of a civil service is slight, -.056. However, mayor-council cities are more often found in the northeast and middle west. The correlation of political form and a sunbelt location is -.449. As we have seen, the relationship between council-manager cities and increased minority workforce representation may well represent factors other than structure. However, the notion of an autonomous bureau as an independent impediment is maintained. The presence of a minority mayor and/or city manager appears to have no impact on minority hiring, judging from the Pearson Correlation. When economic controls are applied, a relationship in the hypothesized direc- tion does emerge, although it is not a strong one. 69 Similar correlations were estimated for the police and fire components of city workforces. These appear in Table 3-12. Table 3-12 Correlations between Minority Representation in the Protective Services and Key Local Determinants Z Minority in Control Control Control Population - for for for Z Minority in Unemployment Workforce both Prot. Services Rate Change Political Form .099 .043 .078 .037 Bureau Autonomy .063 .059 .071 .064 Minority Executive .162 .137 .151 .134 In looking at the results for the protective services component, the negative effects of a mayor-council form or the presence of a civil service commission again appear. With controls for economic factors, the effect of political form almost disappears. The presence of a civil service commission, however, does not lose its negative influence when controls are applied but its effect here is more moderate than for the workforce as a whole. The presence of a minority executive is again found to have a negative effect on minority representation in the protective services though this effect is somewhat attenuated by the application of economic controls (going from .162 to .134). Of course, the possible negative effects of strong police and fire unions with collective bargaining powers on workforce integration cannot be captured at this time. Suffice 70 it to say that these are both bureaus which have insisted on their own professional determination of competency and definition of eligibility requirements. In fact, they sometimes have their own separate civil services (e.g., Detroit, Milwaukee, Denver). Further correlations were estimated for the workforce minus the protective services component. These results are found in Table 3-13. Table 3-13 Correlations between Minority Workforce Representation Excluding Police and Fire and Key Local Determinants Z Minority in Control Control Control Population - for for for Z Minority in Unemployment Workforce both Workforce Rate Change Political Form .164 .049 .136 .046 Bureau Autonomy .176 .161 .171 .161 Minority Executive -.023 -.089 -.041 -.091 When protective services is excluded from the workforce, the correlation between presence of a civil service commission (bureau autonomy) and a shortfall in minority representation rises slightly. Also, the beneficial effect of a minority mayor and/or manager is marginally higher. And, when protective services are excluded, the advantage previously enjoyed by council-manager cities grows even slighter after the application of economic controls. Again, the negative effects of the presence of a civil service commission are barely altered when economic factors are considered and the positive benefits of a minority executive increase. 71 A further set of correlations was estimated to answer a question raised earlier. Aside from differences in structure, council-manager cities are often smaller in both population and minority population than their mayor-council counterparts. To ascertain if these factors may alter the negative relationship between the mayor-council form and minority representation, controls have been applied for both city size and percent minority population. The results are found in Table 3-14. Table 3-14 Minority Representation and Political Form: Controlling for City Size and Minority Population Control for Control Workforce Control for Minority for Shortfall Size Population both POLITICAL FORM Total .203 .185 .167 .160 Workforce Protective serdces 0099 0061 0014 -0001 Without Protective Services .164 .141 .153 .138 In all sets of correlations in Table 3-14, the negative effect of a mayor-council form is reduced by controlling for both city size and minority population. This is a further indication that governmental structure may actually be irrelevent to the question at hand when sufficient controls are applied, contrary to both original expectations and initial findings from analysis of variance. 72 With some confirmation now of the possible effect of city size on results, this variable will be added to the formal model which was speci- fied in Chapter 11. Minority population will be added as well with the stipulation here that its effect on minority hiring could be negative for two reasons. First, the tautologically positive effects of minority population on minority hiring have been eliminated because the dependent variable includes minority population in determining achievement of federal goals. Hence, the strong positive effect found by Eisinger should be greatly diminished and two other circumstances will be given more play. For one, cities with large minority populations are often those cities most beset by economic difficulties. And, these cities too are usually part of the industrial heartland and heavily unionized, particularly in regard to skilled labor and often police and fire as well. The effects of the latter set of circumstances cannot be controlled at this time. Finally, cities with considerable minority populations may have had to adjust their hiring practices to a greater extent to achieve much higher targets. Initial statistical analysis has shown that the importance hypothe- sized to political form may have been exaggerated or even incorrectly directed. However, support has been found for the conception that an autonomous civil service will negatively affect the achievement of a representative workforce. The presence of a minority mayor and/or city manager may have only a minimal effect on minority hiring. However, the nature of its effect appears to be conditioned by nonpolitical and non- organizational factors. The best test, however, of these original and modified hypotheses lies in the estimation of a causal model and it is to this that we now turn. 1. 2. 3. 4. Chapter 3 Footnotes I am not using the .05 level as an automatic cutoff point for findings of statistical significance. Rather, I would like to make mention of relationships where there is a 60Z or greater likelihood that differences in this sample were not caused by chance. This is done in order to earmark hypotheses deemed worthy of continued study. Of course, it must be acknowledged that various other variables might significantly affect these relationships in either direction. For a general discussion, see Ramon E. Henkel, Tests of Significance (Beverly Hills: Sage Publications, 1976), pp. 78-85. Dennis R. Judd, The Politics of American Cities (Boston: Little, Brown and Company, 1979), p. 112. Ibid., p. 113. Ibid. 73 Chapter 4 Modeling Minority Workforce Representation: Expectations and Realities Initial statistical analysis has revealed that two local factors judged as important to successful affirmative action, political form and presence of a minority executive, may in fact have little influence, and that often in the direction opposite to what has been hypothesized. However, those cities without formal civil service commissions did achieve better records of minority representation in every breakdown attempted, in congruence with expectations. Also, economic factors seemed to have a strong impact on minority hiring in the partial correla- tions. These early tests, though, did not utilize all elements envi- sioned for a formal model of minority representation in municipal work- forces. 80, although certain expectations are now somewhat tempered, the model will be estimated as previously envisioned, with two additional independent variables: 74 75 Z Minority in City - Political + Bureau + Minority Population minus Form Autonomy Executive Z Minority in City (O-manager) (O-no comm.) (O-white) Workforce (1-mayor) (l-comm.) (l-minority) + Unemployment + Workforce Rate Change + Residency + Enforcement Requirement (O-no suits) (O-no reqt.) (l-suit filed) (l-reqt.) + City Size + Z Minority + Sunbelt Location Population (O-no) (l-yes) City size and the percent minority in the city's population have been added as a result of correlational analysis discussed in Chapter 3. In order to lessen the possibility of specification error, an algebraic transformation will be applied to the above equation. Percent minority population, as an independent variable, could be correlated to the error term as it is also used to compute the dependent variable. Hence, minority population will be subtracted from each side of the equation. Before reporting on the estimation of this model for the total workforce, the expected effect of each exogenous variable will be reviewed briefly: 1. Political Form - All other things being equal, presence of a mayor-council form of government should increase minority representa- tion. 2. Bureau Autonomy - The presence of a formal civil service commission should impede minority hiring. 3. Minority Executive - If a city has a minority elected mayor and/or city manager, minority represen- tation should increase. 76 4. Unemployment Rate - High unemployment will have a negative impact. 5. Workforce Change - If a city has increased its workforce between 1977 and 1982, minority hiring should increase. 6. Residency Requirement - If a city only hires from within its borders, the opportunities for minority hiring should rise. 7. Enforcement - If a city has been sued because of alleged hiring discrimination, this should stimulate affirmative action effort. 8. City Size -.A smaller city may find it easier to move toward the federal target. 9. Z Minority Population - The greater the minority population, the more difficult the target is to achieve. Also, cities with large minority populations are generally more beset with economic woes and with strong professional unions opposed to affirmative action. A negative effect is expected. 10. Sunbelt Location - A sunbelt location should be favorable to increasing minority workforce representation. This model was estimated, for the total workforce, with Ordinary Least Squares. Tests were carried out to see if the basic assumptions of the general linear model pertained. For this sample, no evidence of heteroskedasticity was discovered. However, multicollinearity was discovered among the independent variables. It is possible that coeffi- cient estimates may be imprecise and standard errors may be inflated. Coefficients may appear to be statistically insignificant when in fact they are not. After examining Farrar-Glauber tests, it was decided that the multicollinearity was not severe enough to threaten the model's 77 utility. Deleting exogenous factors might attenuate the multicollinear- ity but would at the same time increase the danger of specification error. The results of the model estimation are located in Table 4-1. Please note that here again a positive coefficient connotes a negative effect on minority workforce representation and conversely a negative coefficient is indicative of a positive effect. In this regression model, where a number of outside factors could be held constant, all three key local political and organizational variables (political form, minority executive, bureau autonomy) act on minority employment in the hypothesized direction. Only political form, i.e., presence of a mayor-council form of government, is not significant at least the .05 level. The initial statistical findings which placed council-manager cities at an advantage are reversed when all economic and demographic controls are applied. The presence of a minority executive is a significant positive spur to achievement of a workforce representa- tive of population while the presence of a formal civil service commis- sion has an expected negative impact. The various economic variables-unemployment rate, workforce change, sunbelt location-have the hypothesized effects on minority representa- tion in municipal workforces. However, the effect of workforce change is not particularly significant. It might be noted that in the time period under consideration, 1977-1982, a majority of cities experienced some decline in the number of fulltime employees. Contrary to Eisinger's findings, the presence of a residency requirement does further minority workforce representation in a 78 Model of Minority Representation in the Total Workforce Table 4-1 VARIABLE COEFFICIENT Political -1.77 Form (1092) (O-manager) (l-mayor) Bureau 3.01 Autonomy (1.69) (O-no comm.) (1-comm.) Minority -7.11 Executive (2.82) (O-no) (l-yeS) Enforcement 2.29 (O-no suits) (1.98) (l-suits) Unemployment .65 Rate 1982 (.24) City Size -.04—05 (.12-05) Workforce -1.01 Change (3.66) (1977-1982) Residency -3.33 Requirement (1.89) (O-no) (l-yes) Z Minority .19 Population (.06) Sunbelt Location -6.94 (O-no) (2-28) (l-yeS) Constant -4.49 R - .53 R2 . .29 Adjusted R2 - .23 RMSE - 9006 t -092 1.78 -2052 1.15 2.69 -037 -lo77 3.07 -3005 SIGNIFICANCE LEVEL .18 .04 .01 .13 .01 .36 .04 .001 .001 79 significant manner. (Of course, the sample size here is greater, 134 cities as compared to 43.) However, it appears that, contrary to expec- tations, larger cities possess an advantage over small cities in regard to minority hiring but the differences are not statistically significant. On the other hand, the results do show that a larger minority population is negatively related to the achievement of federal hiring goals. This relationship is significant at the .001 level. Finally, the expected positive effect of court suits on minority hiring is not present. Rather, court suits appear to be related to larger shortfalls between goals and actual representation. It would be specious to assume that this is causal. Rather, it appears that those cities which have been sued are often those with the poorest hiring records. Perhaps, the time frame, 1973-1981, is insufficient to adequately judge the effect of these suits. In addition, not all have been adjudicated at this writing and it is also possible that disposi- tions in some suits which were decided in favor of the plaintiffs did not specify clear timetables for action in regard to minority hiring. As a whole, the model explains 23Z of the variation in the dependent variable. A number of critical variables do act in the manner hypothe- sized and are significant at the .05 level. It is readily acknowledged that certain factors, which will be explored in greater depth in Chapter 5, have not been included: political and bureaucratic attitudes toward change, actual modifications in hiring practice, the political base of key elected officials, the attitude of various municipal unions. These factors will all need more detailed study in order to fully capture the processes at work which influence achievement of a city workforce 80 representative of city population. With this caveat, one can say that progress has been made in understanding local political and organiza- tional dimensions at work as well as their relationship with the economic health of the local community. It has been my contention that an appropriate dependent variable is one which measures the number of minority employees against some sort of implementation target. Previous research in this area, notably that of Peter Eisinger, has not done so and has purported to achieve more power- ful explanations of variance. It is thus useful to re-estimate the model employed here utilizing the dependent variable used in Eisinger's study, namely percent of minority employees. I assume that the adjusted R2 will be elevated and that minority population, on the right side of the equation, will provide a great deal of the enhanced explanatory power. Table 4-2 contains the results of this re-estimation, alongside the results of the model with the dependent variable based on achievement of mandated goals. (It should be noted again that the sample size is a good bit larger than in the Peter Eisinger study and that I have included all protected minorities rather than blacks alone.) As Table 4-2 amply illustrates, the expected results of the compara- tive examination have been achieved. As both dependent variables are in the same metric, analysis is simplified. Except for one variable and the goodness of fit measures, the results of both models are identical. There are identical coefficients and levels of significance for all variables save minority population. The rest of the dependent variables influence the percent of minority workers in the same fashion as they impact on goal achievement. However, their contribution to explanation of variance does differ between the models. 81 Table 4-2 Two Models of Minority Workforce Representation: A Comparative Assessment With Z Minority in Workforce as Dependent Variable (positive coefficient indicates positive effect on Z minority in workforce) With Z Minority in Population - Z Minority in Workforce as Dependent Variable (positive coefficient indicates a negative effect on work- force representative of population) VARIABLE b t level b t level Form (1.92) (1.92) (O-manager) (l-mayor) Bureau -3.01 -1.78 .04 3.01 1.78 .04 Autonomy (1.69) (1.69) (0-no comm.) (l-comm.) Minority 7.11 2.52 .01 -7.11 -2.52 .01 Executive (2.82) (2.82) (O-no) (l-yeS) Enforcemnt -2029 -1016 013 2.29 1.16 013 (O-no suit) (1.98) (1.98) (l-suit) Unemployment -065 -2069 .01 .65 2.69 001 Rate 1982 (.24) (.24) City Size 04-06 .37 036 .04-06 "037 036 (01"05) (cl-'05) Workforce 1.01 .28 .39 -1.01 -.28 .39 Change (3.66) (3.66) 1977-82 Regidency 3033 1.77 004 -3033 -1077 .04 Requirement (1.89) (1.89) (O-no) (l-yes) Z MINORITY .81 12.90 .001 .19 3.07 .001 POPULATION (.06) (.06) Sunbelt 6094 3005 .001 -6094 -3005 0001 Location (2.28) (2.28) (O-no) (l-yes) Constant 4.49 -4.49 R2- .87 R2- 053 Adjusted R a .73 Adjusted R a .23 82 When the dependent variable is not tied to a target figure, the coefficient for the percentage of minorities in a city's population greatly increases as does its t ratio. In fact, the minority population variable accounts for 44Z of the R2 change in the reestimated model. An adjusted R2 of .73 is achieved as compared to the .23 for the model which ties workforce representation to population representation. If one wishes to explain which cities will have a higher percentage of minority workers, demography, i.e., the percentage of minorities in a city's population, is clearly the best predictor. In this comparison, and presumably in Eisinger's early models, the significance of other exogenous factors is not disturbed. However, this type of model cannot be successfully utilized if one is interested in the achievement of policy objectives. The demographic explanation would make one confident of one's results but would not explain the relative success of a city with a lSZ minority population as opposed to a city with a 60Z minority population. Except for very unusual circumstances, the city with the larger minority population would generally have a greater minority component in its workforce. One could not generally expect a city with a 15Z minority population to far exceed that figure in workforce represen- tation. So, with an untargeted dependent variable, only cities with large minority populations could be successful. In order to examine city hiring practices in a policy context, cities must be ranked by their relation to the target that has been specified in various orders issued by the federal government. In that way, some of the conditions related to compliance can be highlighted and the fact that there is more to be learned is also made clear. Of course, 83 the selection of an appropriate model must be contingent upon the question being asked. Eisinger did not place his question in the type of policy context that has been laid here. As such, his work offered guide- posts for future study and highlighted important variables to be included in the type of revised model utilized here. It should be noted though that by changing the dependent variable to fit the context of affirmative action policy, the explanatory power of the results is reduced, as is the probability of Type II error. In fact, in a work published very recently by the Joint Center for Political Studies, Eisinger himself has abandoned his previous approach.l However, here too, he has not utilized a policy target and his sample is again quite small. He looked at 40 cities with populations over 50,000. In this work his dependent variable is the change in the percent black in the total city workforce between 1973 and 1980. Of his eight independent variables, four are statistically significant: percent black in city in 1980, black mayor 1979, growth of per capita revenue 1973-1980, growth of per capita state aid 1973-1980. The latter variable, however, is significant opposite to the direction hypothesized. Eisinger found that the greater the state aid to a city, the smaller is the increase in black hiring. The R2 for this model is .63. (An adjusted R2 is not reported.) The model is more sophisticated than those of his earlier studies but again the results are dependent on demography. If a city has a 15Z black population, a 3Z increase may have allowed it to reach a workforce representative of city population while a city which has a 53Z black population might fall short even after a 9Z gain. 84 Although recent actions of the Reagan administration cast some doubt on the following words, it has been the policy of the U.S. government to require that the nation's cities achieve workforces representative of their pOpulations. There is a discernible target by which to gauge compliance with federal orders. To measure the process at work in American cities, that target must be included. Finding only 23Z of the variance explained in the model using a dependent variable tied to a policy target is not a liability. The strengths and weaknesses of political and organizational factors are revealed and the influence of demography is seen in a far different light. Regarding sheer workforce composition, we can make substantial knowledge claims. Regarding compliance with a federal administrative order, we can say progress in understanding the process at work has been made but it is clear that further research is desirable. The measurement of the dependent variable is critical in understanding what is and is not explained. Protective Services Results from initial statistical tests showed that the police and fire departments of municipal workforces differ from the total workforce in the degree to which they approach representation based on city popula- tion. There are greater shortfalls in the protective services. Also, the three local political and organizational factors affect minority representation in the protective services to a lesser degree or, in the 85 case of a minority mayor and/or manager in a reverse direction. Controls for economic factors did exercise a tempered effect on all three factors. The model to be used now to explain minority employment in terms of the federal goal is identical to that used previously for the total work- force and the hypothesized effects of the various variables remain the same. However, based on the initial tests, the results are not expected to be identical. They are located in Table 4-3. The results in Table 4-3 show that for the model of minority repre- sentation in the protective services, the three key local factors act in the hypothesized manner. However, the presence of an autonomous civil service commission does not have the statistically significant negative effect it had for the total workforce. But, with controls for economic factors, demographics and enforcement, both the presence of a mayor-coun- cil form of government and a minority executive achieve significant positive effects at the .05 level. The earlier tests which portrayed minority-governed mayor-council cities at a considerable disadvantage are completely reversed. This of course demonstrates again that the various factors used in this model are interrelated and that either a solely political or a solely economic explanation would produce misleading results. With regard to the economic variables, the unemployment rate and sunbelt location are statistically significant in the expected direction. However, workforce change has a negative effect on representation, implying that cities which increased their number of employees between 1977 and 1982 had less success in adding minorities to the police and Table 4-3 Model of Minority Representation in the Protective Services Compared to Model of the Total Workforce PROTECTIVE SERVICES VARIABLE b t level Form (2.00) (O-manager) (l-mayor) Bureau 1.44 .82 .21 Autonomy (1.78) (O-no comm.) (l-comm.) Executive (2.93) (O-no) (l-yeS) Enforcement 6.36 3.11 .001 (O-no suit) (2.04) (l-suit) Unemployment .18 .72 .24 Rate 1982 (.25) City 8128 01-06 .08 .47 Workforce 2.03 .53 .30 Change (3.82) 1977-82 Requirement (1.95) (O-no) (l-yeS) Z Minority .33 5.08 .001 Population (.07) Sunbelt -4.22 -1.79 .04 Location (2.36) (O-no) (l-yes) Constant 1.23 R an .57 R2 ' .32 Adjusted R2 a .26 RMSE - 9.38 TOTAL WORKFORCE b t level -1077 092 018 (1.92) 3.01 1.78 .04 (1.69) -7011 -2052 .01 (2.82) 2.29 1.16 .13 (1.98) .65 2.69 .01 (.24) -04-06 -o37 036 (01-05) -1001 '028 039 (3.66) -3033 -lo77 .04 (1.89) .19 3.07 .001 (.06) -6094 -3005 .001 (2.28) '4049 R = .53 R2 - .29 Adjusted R2 3 023 RMSE 3 9.06 87 fire departments. This is contrary both to expectations and to the results from the model of the total workforce. Of course, it has also been posited that police and fire departments do represent separate patterns; the correlation between achievement of a representative work- force in the protective services and in the workforce as a whole is only .31. Finally, it is possible that whatever workforce cutbacks did occur were not evenly distributed among various city departments. A residency requirement has a positive effect on increasing minority workforce representation but this in effect is less significant than it was for the total workforce. In this model, enforcement again has a strong negative effect on minority hiring. This must be viewed, for a second time, as correlative rather than causal. The majority of suits brought by the federal government and private parties have involved the protective services. Some of the suits have not yet been adjudicated and it is possible that for those which have insufficient time has passed in which to see the substantive results of legal action. It is appropriate here to make mention of the actions of the Reagan administration with regard to enforcement. Birmingham, Alabama had entered into a consent decree in 1981 with regard to municipal hiring as a result of action by the U.S. Department of Justice. The city agreed to take action to increase the number of minority police officers. However, in February of 1984 white officers filed suit and charged reverse discrimination. Here, the Justice Department has entered the case--on the side of the white officers-though it was this very department with the same Attorney-General which participated in the formulation of the consent decree. 88 Further, the Memphis fire department experienced layoffs in 1981. A black fire captain requested that Memphis' consent decree be amended in order that recently hired black firefighters be protected from dismissal. The district court judge concurred as did the U.S. Circuit Court. This case went before the U.S. Supreme Court with the Reagan administration providing assistance to the white firefighters who protested the amend- ment of the consent decree. In June 1984, the Supreme Court did rule in their favor. In another recent case, the Reagan Justice Department took the side of the Detroit Police Officers Association when it appealed a promotional plan for the Detroit Police Department. The plan called for a black officer to be promoted each time a white officer was elevated in rank. The Detroit Police Department historically has had very few minorities above the rank of patrolman. In this instance, the Supreme Court refused to hear the case, letting stand lower court rulings in support of the promotional plan. Two important ideas emerge from this examination of recent events. One is that whatever consistency the federal enforcement effort may have had in the 19703 has dissipated since January 1981, as has the verbal commitment to affirmative action to address past workforce discrimina- tion. Further study is needed in this area, particularly regarding the enforcement of consent decrees stemming from earlier suits. Another idea is that mentioned in the preceding chapter: the possible negative effects of protective services unions with collective bargaining powers. Regrettably, these cannot be measured at the present time but they should be legitimate objects of further inquiry. 89 Regarding demographics, the results of the protective services model show that city size has a weak effect on achievement of a representative workforce. However, for the protective services, larger cities are associated with less minority representation in terms of population. For the protective services as for the total workforce, the percent of minority population in a city is associated with greater shortfalls in goal achievement. As the protective services were the least integrated bureaus initially, the increases needed to achieve representation in terms of population are generally greater than for the workforce as a whole and of course greater for cities with large minority populations. Resistance to the changes needed to facilitate affirmative action in this case would occur not only from personnel professionals but also from police and fire professionals who have enjoyed considerable autonomy in many locales and who have also resisted political direction. An example of the professional determination of professional prac- tice in the protective services can be found in Lowi's work on New York City. Lowi found that the mayor always had to appoint a law enforcement specialist as Police Commissioner, someone from the New York Police Department or the New York District Attorney's office.3 A nontradi- tional appointment would not be tolerated. In a more general vein, James Q. Wilson has noted that "deliberate community choices rarely have more than a limited effect on police behavior.”4 Police department budgets, staffing, and organization can be subject to mayoral or managerial action. But, the operating proce- dures and objectives of the organization ”are, with very few exceptions, determined by the police themselves without any deliberate or systematic 90 intervention by political authorities."5 With regard to relations with the minority community, Wilson found that police officers generally saw this area as one of political challenge to their authority and as a threat to their career interests including salaries and promotions.6 He concludes by saying that police bureaus are more resistant to change than most and that resistance will be most manifest when complaints, grievances or conditions of employment are involved.7 Resistance to a process of civilian review by police officers is at least partially based on the fact that non-police professionals would be judging police perfor- mance. Therefore, based on the literature, it would not be surprising for autonomous police and fire departments, particularly in cities with large minority populations, to view affirmative action as a threat to personal security, a traditional base of power, and to established prac- tices. In the protective services model, the adjusted R2 is slightly higher than for the workforce as a whole (.26 to .23). However, this result may be due in part to the increased strength of the enforcement coefficient and hence be misleading. The slightly higher root mean square error shows that the model does not fit the protective services data as closely. ' The model's importance lies in the fact that it shows that political variables do perform in the manner expected when other exogenous factors, principally economic, are taken into account. For the protective services, as for the total workforce, cities in high growth areas with low unemployment fare better. Growth and unemployment in turn greatly affect other hypothesized relationships, especially dealing with govern- ment structure and the presence of a minority mayor and/or manager. 91 As the protective services model does differ from that of the total workforce, it will again be interesting to examine the workforce when protective services is deleted. Modeling the Workforce without Protective Services Based on examination of initial statistical tests, it is expected that a model of the workforce without protective services would most closely resemble that of the workforce as a whole. The expectations for the coefficients are those posited at the beginning of this chapter. Model results are located in Table 4-4. The model presented in Table 4-4 is indeed similar to the model of the total workforce. The coefficients representing local political and organizational factors are in the expected direction and of about the same magnitude. Economic variables, particularly the unemployment rate and sunbelt location, continue to provide considerable explanatory power. However, workforce change, as in the protective services model, has a surprising negative effect on minority workforce representation. In another difference, the insignificant effect of city size is now posi- tive; the larger the city, the more likely it is to achieve a workforce representative of population. In this aspect, the model resembles that for the total workforce rather than the protective services model. When the police and fire units are removed from city workforces, the enforcement variable at last achieves its expected positive direction although it is not significant. It is possible that for those cities 92 Table 4-4 Model of Minority Representation in the Workforce with Protective Services Excluded Compared to Model of the Total Workforce EXCLUDED VARIABLE b t Palitical -1095 ”.83 Form (2.33) (O-manager) (l-mayor) Bureau 4.64 2.26 Autonomy (2.06) (O-no comm.) (1-comm.) Minority -6.58 -1.91 Executive (3.44) (O-no) (l-yeS) Enforcement -1.01 -.42 (O-no suit) (2.42) (1-suit) Unemployment .80 2.72 Rate 1982 (.29) City Size -.4-06 -.29 (02'06) Workforce .66 .15 Change (4.45) 1977-82 Residency -5.11 -2.22 Requirement (2.30) (O-none) (l-yes) Z Minority .18 2.37 Population (.08) Sunbelt -9o37 ’3038 Location (2.77) (O-no) (l-yes) Constant -9.70 R a .50 R2 . .25 PROTECTIVE SERVICES Adjusted R2 e .19 RMSE - 11.04 level .20 .02 .03 .34 .01 .39 .44 .01 .01 .001 TOTAL WORKFORCE b t level -ls77 092 .18 (1.92) 3.01 1.78 .04 (1.69) "7.11 -2052 001 (2.82) 2.29 1.16 .13 (1.98) .65 2.69 .01 (.24) _04-06 ”.37 .36 (01"05) -1001 "028 .39 (3.66) -3033 -lo77 004 (1.89) .19 3.07 .001 (.06) -6094 -3005 .001 (2.28) -4049 R a .53 R2 3 029 Adjusted R2 3 .23 RMSE 8 9.06 93 which experienced employment-related litigation, increases in minority hiring occurred in units outside the protective services even when the police or fire department was a party to the suit. This model does not fit the data as well as its two previous counterparts. Only 19Z of the variance has been explained and the Root Mean Square Error is larger. An Assessment of the Results The models discussed above provide considerable insight into the problem of achieving a municipal workforce representative of population. Local political and organizational variables-political form, bureau- cratic autonomy, presence of a minority executive--act in the hypothe- sized manner, when all economic and demographic controls are present. Of particular interest is the continuing negative effect of a formal civil service commission. This lends weight to the argument that an autono- mous, professional bureau has more ability to resist change to existing practices, particularly change emanating from outside the profession. The strong impact of economic factors on a variety of political out- puts found in studies at the state and local level is again demonstrated in this work. Cities experiencing more growth and less unemployment were far more able to increase the number of minorities on their payrolls. Economics is not the sole explanation here but its impact is strong. In fact, economic health is the major reason for the greater goal achieve- ment of council-manager cities witnessed in the initial breakdowns. 94 Finally, we have seen that the introduction of a goal-specific dependent variable can greatly alter a model's results and is necessary to understand workforce representation within a policy context. It is an interesting exercise to see which cities have larger percentages of minorities in their workforces. But, if minority representation is not related to population on the left side of the equation, no insight is gained regarding compliance with federal administrative orders which set a specific goal for the nation's cities. Chapter 4 Footnotes Peter K. Eisinger, Black Employment in City Government, 1973-1980 (Washington, D.C.: Joint Center for Political Studies, 1983). Ibid., p. 43. Lowi, At the Pleasure of the Mayor, p. 159. James Q. Wilson, Varieties of Police Behavior (Cambridge: Harvard University Press, 1978), p. 227. Ibid., pp. 227, 230-231. Ibid., p. 231. Ibid. , pp. 231-232. 95 Chapter 5 Minority Municipal Workforce Participation: Toward Explaining the Unexplained The models of minority workforce representation, explicated in the preceding chapter, have demonstrated the significance of local political, organizational, economic and demographic factors. It has been shown that a political form generally judged to be more receptive to the needs of minorities and the disadvantaged is indeed more receptive when economic and demographic controls are present. A city with an elected chief exec- utive with ultimate administrative authority is more likely to utilize public employment for the purposes of affirmative action. Similarly, if the executive, elected or not, is a member of a minority group, there is greater likelihood of increased minority hiring. Also in congruence with expectations, the presence of an autonomous civil service commission has been seen to have a negative effect on achievement of a workforce repre- sentative of pOpulation. Finally, it has been found that local economic conditions as well as demography affect workforce composition in a signi- ficant manner and also affect the behavior of political variables. Rather than nullifying political effects, controlling for economic and demographic conditions allows differences due to political form and the presence of a minority executive to manifest themselves. The results from the models lend considerable support to the key hypotheses of this research. However, certain questions remain to be 96 97 answered. For one, why has minority hiring occurred to a much greater degree than anticipated in the total workforce? Also, why does minority hiring in the protective services lag far behind hiring in the workforce as a whole in almost every city studied? It is important to ask at this juncture if there are ways to build on the factors which have proven significant in order to explain a greater degree of the variation in minority representation. Political, organizational, economic and demo- graphic variables have demonstrated their utility in explaining minority representation in municipal workforces. It is possible to enhance the significance of these factors through refinement and through utilization of different forms of measurement. The model results have demonstrated the value of the type of approach undertaken and contain the seeds of additional understanding. In the pages that follow, possible elaborations of the significant explanatory factors will be discussed in order to demonstrate how further study in congruence with the basic premises of this research may fine tune the findings to date and explain a greater portion of the variation in minority representation. Hiring Minority Workers: An Allocational Strategy? In this study of employee composition in 134 cities, it was discovered, with some surprise, that the average shortfall in achievement of a workforce representative of population was very slight, 2.18Z. It had been hypothesized that implementation of the federally-mandated 98 affirmative action policy would encounter significant resistance. Certainly, a number of cities do have considerable deficits when examined in terms of the federal goal. But many--in all areas of the country- have come close to or have exceeded the target set for them. Because of this relative success, it is worthwhile to discuss a theory advanced by Paul E. Peterson. In an important work published in 1981,1 Peterson specified three fundamental urban policy areas: devel- opment, allocation, redistribution. Because of a limited ability to generate revenue, developmental strategies designed to increase the urban tax base would be eagerly pursued. For the same reason, circumscribed resources, special redistributive policies aimed at the poor and power- less would be eschewed. On the other hand, allocational policy, involv- ing such items as street repair, location of parks or fire stations, and garbage collection schedules, deals with dispersal of resources to one area of the city or another and can be adjusted for political reasons. Peterson feels that municipal employment is also a clear example of allo- cational policy and hence subject to manipulation. In employment, as in other allocational areas, size and expenditures can be contained but direction can be changed without additional monetary cost. An alloca- tional strategy can be employed to serve political ends without causing harm to a city's precarious fiscal position. The civil rights demonstrations and urban riots of the 19603 led to a recognition by mayors and city managers that steps had to be taken to soothe the waters and quell disturbances. Substantive redistribution by local officials was out of the question. However, municipal employment, according to Peterson, was seen as an area where adjustments could be made.2 He notes that, 99 Community action programs, Model Cities programs, manpower and training programs, and community health centers were staffed primarily with individuals from minority backgrounds. At the same time affirmative action programs were undertaken by old-line governmental agencies. Although bureaucratic resistance to minority appointments was often considerable, government leaders gradually intrqduced new procedures with less obvious racial biases. The empirical evidence Peterson offers, however, is not fully convincing. He footnotes Frank J. Thompson's case study of the Oakland personnel department4 as substantiation. It is true that, in Oakland, the mayor and city manager responded to black demonstrations in 1969 by attempting to increase the hiring of blacks. However, they limited their efforts to one category: firefighter. Thompson documents that the changes in minimum requirements and to the written examination for the position were very minor. Further change was successfully resisted by personnel administrators and fire department officials. Only a few blacks were actually hired.5 Secondly, Peterson has cited a study that he and David Greenstone conducted which compared the employment of blacks in local government with black employment in the U.S. as a whole. As Peterson and Greenstone found a higher proportion of black employees employed by local government in various categories, they concluded that blacks were overrepresented on the local payroll.6 What was not asked was whether it may have a question of black underrepresentation in the economy as a whole. Perhaps private discrimination is even more enduring than public. Although Peterson's empirical evidence may be faulty, his general premise may not be. In cities with limited resources, increasing access to municipal jobs may indeed be a way of coping with minority demands. 100 In order to test Peterson's theory and to account for the high degree of minority representation this research has found, it is neces- sary to further explore the attitudes of mayors and city managers toward minority hiring. Is the allocational strategy a conscious answer to various minority protests? A possible line of study would be to corre- late instances of unrest with hiring patterns. A further question asserts itself here. Minority unrest may also be coupled with minority political strength. If the number of minority voters is growing, an allocational strategy might be adopted to increase political support. In that vein, it would be necessary to examine the political base of support that various elected mayors enjoy. Those that include minority citizens in their coalition may be more likely to attempt to affect job allocations. It is important to note here that after economic and demographic controls were applied, the mayor-council form of government proved more amenable to an increased minority presence in the municipal workforce. As Thompson has noted, minority groups may well focus attention on the area of municipal jobs as ”this is the arena where their efforts bear the most fruit."7 Opposition may be less than in other areas, such as the desire to create civilian review boards for police forces. And, ”minority advocates. . . sense that black officials will do more to represent minority interests."8 In reference to this point, it would be worthwhile to examine which cities had experienced direct minority pressure in regard to municipal hiring. Peterson's allocational strategy provides certain political reasons for increasing minority municipal employment. Given the fact that we 101 have learned of the positive influence of both the mayor-council form of government and the presence of a minority executive on minority hiring, this strategy provides an additional means to ascertain political deter- minants. To more fully comprehend the political dimension, executive attitudes, political base and minority demands must also be measured. Before beginning this project, I downplayed Peterson's notion of the receptivity of high city officials to increased minority hiring. However, after seeing the level of minority representation in the sample cities, the idea appears more likely. Yet, the obstacles to the desires of mayor and city manager that Thompson documented in Oakland could still circumscribe the actual utilization of an allocational strategy. The Organizational Context: Dimensions of Professionalism and Authority The Personnel Profession In this research, the presence of a formal civil service commission has been shown to have a negative effect on the attainment of a workforce reflective of city population. It had been hypothesized that such a commission would prove more resistant to the changes needed to implement an affirmative action program and that its autonomy would enhance its capacity to resist. Statistical tests have corroborated this argument. Yet, to fully understand the actions of municipal personnel profession- als, further inquiry along two related dimensions would be desirable. 102 The first dimension would involve an examination of the attitudes of personnel administrators in various cities toward affirmative action policy and the procedural changes it necessitates. Attitudes could then be compared to actual changes in the nature of professional practice and the degree of minority representation achieved. In their study of school desegregation in Georgia, Rodgers and Bullock found that compliance with federal court decisions was affected by "whether the regulated agree with the legal standard [and] whether the regulated feel they would benefit from the law."9 Considerable groundwork has been laid in the study of attitudes of personnel professionals. In 1975, Frank J. Thompson surveyed members of the International Personnel Management Association who worked for some form of locally-based government (state, city, county, special district). Thompson received almost 1000 responses. Unfortunately, for purposes of this work, data were not broken down by city. However, some of the findings based on this survey are interesting and illustrate the possi- bilities for additional research. In a 1978 article, Thompson and Brown reported that commitment to the merit principle on the part of personnel professionals was negatively 10 Adherence to correlated (-.22) to a commitment to minority hiring. the precept that the most competent people available should be hired is indicative of less support for the active recruitment of minorities. In a further work based on his survey, Thompson reported that 40Z of resondents did agree that "public agencies should establish hiring targets for minorities and timetables for obtaining them."11 A larger percentage, 71Z, agreed with the statement that "where minority 103 and white applicants are of about equal ability and minorities are under- represented on the department's work force, public agencies should give preference to minority applicants."12 These responses indicate a certain degree of support for affirmative action principles. However, it must be noted that survey respondents may work at personnel departments as well as formal civil service commissions. A further set of correla- tions (located in Table 5-1) examine the relationship between receptivity to hiring minorities and measurements of professionalism. Though these correlations are positive, they are generally quite weak. Table 5-1 Correlations between Receptivity to Minority Hiring and Measures of Professionalism* Receptivity to Hiring Minorities (r) Fulltime in personnel .17 Greater education .10 Graduate specialization in personnel .09 More frequent attendance at professional .06 meetings Belief in reading professional journals .06 Commitment to a career in personnel .01 *adapted from Frank J. Thompson, "Civil Servants and the Deprived: Socio-Political and Occupational Attitudes toward Minority Hiring,” 22, 2 American Journal of Political Science (May 1978), p. 338. The findings of Thompson, and Thompson and Browne demonstrate a relatively high level of support for at least part of an affirmative action program but at the same time indicate that those who rank highest 104 in measures of professionalism have only a minimal amount of sympathy for program goals. Relating attitudinal findings to employment results should increase the negative impact of an autonomous, professional bureau. Yet, relating attitudes to results would be incomplete unless one also assesses concrete adjustments to personnel practices which would benefit minority job applicants. Thompson has found that administrators who are receptive to the increased hiring of minorities are also more likely to favor the redesign of positions so that education and experi- ence requirements can be reduced.13 The question to be asked now is whether such redesign occurred and where. If position requirements were redrawn and recruitment efforts extended, this ought to have a positive impact on minority workforce representation. Naturally, it would be expected that such changes would not occur as rapidly, if at all, in cities with formal civil service commissions. The negative impact of a formal civil service commission found in the formal models makes an interesting addition to the literature on personnel administration and on the role of autonomous bureaus. By adding the two additional levels of study specified here, the role of the professional bureau and bureaucrat can be explicated further and addi- tional variation among cities may be explained as well. The Protective Services Representation of minorities in police and fire departments has lagged behind representation in other segments of the workforce. As has 105 been noted, the majority of federal suits regarding employment discrimi- nation filed during the past twelve years have involved police or fire departments. Part of the reason is historical: police and fire depart- ments had traditionally hired the fewest minorities. Hence, greater distance would have to be travelled to achieve a workforce representative of population. In addition, police and fire departments are exemplars of autonomous, professional bureaus, similar to civil service commissions in their expected resistance to change mandated by those outside their ranks. James Q. Wilson has balked at the word "professional" as an appro- priate appelation for the police officer. Rather, he sees the officer as ”a member of a craft.”14 He feels the craft designation is appropri- ate because, As with most crafts, his has no body of generalized, written knowledge nor a set of detailed prescriptions as to how to behave-it has, in short, neither theory nor rule. Learning in the craft is by apprenticeship, but on the job and not in the academy. The principal group from.which the apprentice wins (or fails to win) respect are his colleagues on the job, not fellow members of a discipline or attentive superiors. And the members of the craft, conscious of having a special skill or task, think of themselves as set apart from society, possessors of an art that can be learned only by experience, andsin need of restrictions on entry into their occupation. The craft versus profession designation may seem less relevent in light of the growing number of community college and university classes in police administration and criminal justice. But, in truth, the signifi- cance of on-the-job learning and dependence on one's fellow officers can- not be disputed. 106 Police work does become a fraternal chore and one can comprehend the desire of officers to admit those on whom they can place trust for their personal safety. To the traditional white occupants of this public defender role, this has at times meant exclusion of minorities. In this aspect, as well as in the on-the-job training area, the police craft bears resemblance to the various building trades unions. Such unions, representing carpenters, electricians, etc., have had a long history of insularity and resistance to integration of their membership.16 The craft designation afforded the police by Wilson can also be applied to firefighters. Their training also takes place on the job and they too are dependent on each other for assistance and protection. While police officers may share patrol cars, firefighters share living quarters and alternate use of beds with their coworkers. Although a literature on firefighters comparable to that on the police has not yet appeared, it is possible to infer that the same insularity and dependence on one's fellows for direction occurs as does resistance to outside interference in working conditions, procedures or conditions of employ- ment. Police officers in many cities over the years have banded together in benevolent associations. In more recent times, a number of these associations have achieved collective bargaining powers and have become, in effect, unions. These associations/unions have been reflective of the desires of white officers, clearly in the majority at the time of organi- zation. Police unions have enjoyed considerable power for two reasons: a high level of political participation and the threat of a strike. 107 Regarding political participation, Wilson has cited examples of a police union publicly supporting candidates for office and staging demon- strations in support of increased compensation and benefits for officers.17 In Philadelphia and Minneapolis police officials successfully sought the mayoralty. Although most local laws limit campaign activity by city employees, this restriction has not adversely affected political participation by those in the protective services, participation designed to protect material well-being and the ability of the police and fire units to determine independently policy for their respective jurisdictions. And, strikes by law enforcement and fire- fighting personnel, once unheard of, have occurred in a number of major cities in recent years. Knowledge of the organizational behavior of protective services bureaus and of the activity of police and fire unions gives rise to the belief that these bureaus and their labor representatives would greet affirmative action measures with recalcitrance and the data bear this out. As police and fire occasionally have a separate civil service, the feasibility of resistance increases. Additionally, police and fire departments often play a more direct role in employee selection than other municipal bureaus. They may review all eligible applicants instead of selecting just from among the top scorers for a position. In Lansing, Michigan, protective services personnel had used the oral interview process to select out females and minorities.18 In order to explain variation in minority representation in the protective services in greater depth, the degree of autonomy of police and fire departments in various cities must be gauged, in general, and in 108 reference to recruitment and selection. Similarly, it will be important to discover which local police associations are in fact unions with collective bargaining power and whether those unions participate actively in local politics. Federal Enforcement Efforts In this work, court suits brought by the federal government or private parties were found to be negatively correlated with workforces representative of population in two of the three models. A likely expla- nation is that suits, in many cases, were filed against those cities with the poorest minority hiring records. Lack of disposition or a too brief passage of time makes this enforcement measure a very weak determinant of affirmative action. Court suits are also an incomplete indicator of federal compliance efforts. Several federal bureaus have proved reluctant to furnish aggre- gate data regarding directives given to specific cities to prepare affirmative action plans or threats to deny funding to cities making insufficient progress toward mandated goals. This reluctance may be due to inability to expend staff time on data compilation or it could be reflective of the attitude of the present administration toward affirma- tive action. In any case, a possible solution to this information block might be a series of select interviews with officials at the EEOC, the Office of Management and Budget and various federal departments which 109 make grants to cities. These interviews would be used to build a more informed history of compliance strategy, practice and success. It was originally hypothesized that the federal role would be neither consistent nor constant. The data generated thus far can neither confirm nor refute that contention. What has been learned is that court suits generally affect those cities with the worst minority hiring records, especially in regard to the protective services. Economics and Demographics The significance of economic factors in explaining variation in minority municipal employment has been reiterated in this research. What may be needed in the future is a refinement of the economic measurements to more closely capture the real world process. Per capita municipal revenue is one possible measure. Whatever the case, it is clear that cities which suffered less in the recent severe recession and which are in high growth areas are more successful at raising their number of minority employees. And, it is also clear that certain political rela- tionships do not become clear until economic differences are taken into account. Regarding demography, there is one significant finding that warrants further investigation. Using a goal-specific dependent variable, it was found that the greater the percentage of minorities in a city's popula- tion, the more difficult it would be to achieve a municipal workforce representative of population. This relationship was statistically 110 significant at the .01 level in the presence of economic, political and organizational controls. Minority hiring in cities with large minority populations must be tied to examination of the role of municipal unions and the attitudes and activities of key actors. In addition, it must be noted that the protective services component generally encompasses one quarter to over one half of the city workforce for cities examined here. The failure of minorities to penetrate this component to the same degree as the rest of the workforce may prove more critical when studying cities with a minority population greater than 502. Further, there may also be an additional means of capturing the severe economic distress experienced by a number of cities with very large minority populations. Another interesting demographic phenomenon worthy of exploration is the very poor performance of several Northeastern cities with sizeable minority populations (30-60Z). These cities have had political machines and it would be interesting to see if minorities had been traditionally outside the base of their machines. In Summation The attempt at modeling minority representation in city workforces in terms of federal policy goals has shown that local political and organizational variables have an impact on results when local economic conditions are taken into account. These findings are in congruence with pertinent literature on urban politics and public administration. Future students of this particular policy area or of other federal policies 111 geared toward minorities or the disadvantaged should make use of these factors. And, greater elaboration should serve to make them more potent predictors of the variation in results. Economic factors have again demonstrated their utility in under- standing policy outcomes. Although a city's growth and economic health significantly influence its rate of minority hiring, they do not stand alone as explanations for achievement of a workforce representative of population. However, they greatly affect the influence of other elements in the equation and should also be elaborated for a more complete expla- nation of variation among cities. It must also be noted that in examination of implementation of or compliance with federal programs or directives, the dependent variable must be goal-specific. A more openended approach such as utilizing percent of minority employees on city payrolls can prove informative but in terms of assessing achievement of specified objectives it is wanting. The method of phrasing a question can often dictate its answer and that cannot be forgotten when attempting to assess local responses to federal mandates. The aim of this study has been to earmark key local determinants which can affect successful implementation of federal administrative orders or programs. To that end, a credible beginning has been made. The variables used here-political, organizational, economic, demo- graphic--have demonstrated their relevance, both intrinsically and as precursors of more sophisticated measures of the same phenomena. These exact variables may not be pertinent to all studies of implementation of federal policy at the local level. But, it is clear that the concepts they represent cannot be excluded from studies of this nature. 1. 2. 3. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Chapter 5 Footnotes Paul E. Peterson, City Limits (Chicago: The University of Chicago Press, 1981). Ibid., pp. 158-162. Ibid., pp. 160-161. Frank J. Thompson, Personnel Policy in the City: The Politics of Jobs in Oakland (Berkeley: University of California Press, 1975). Ibid., pp. 112-139. Peterson, City Limits, p. 161. Thompson, Personnel Policy in the City, p. 169. Ibid., pp. 169-170. Harrell R. Rodgers, Jr. and Charles S. Bullock, III, Coercion to Compliance (Lexington, MA: Lexington Books, 1976), p. 125. Frank J. Thompson and Bonnie Browne, ”Commitment to the Disadvantaged among Urban Administrators: The Case of Minority Hiring,” Urban Affairs Quarterly 13, 3 (March 1978): 365. Frank J. Thompson, ”Civil Servants and the Deprived: Socio- political and Occupational Explanations of Attitudes Toward Minority Hiring,” American Journal of Political Science 22, 2 (May 1978): 330. Ibid., p. 331. Ibid., p. 333. James Q. Wilson, Varieties of Police Behavior, p. 283. Ibid. Benjamin W. Wblkinson, Blacks, Unionsy_and the EEOC (Lexington, MA: Lexington Books, 1973), pp. 9-34. James Q. Wilson, Varieties of Police Behavior, p. 257. Interview with Richard Letts, Human Relations Director and Affirmative Action Officer, City of Lansing, 22 September 1983. 112 APPENDIX Sample Cities and Their Characteristics CITY Akron Albuquerque Alexandria Amarillo Anaheim Anchorage Ann Arbor Atlanta Aurora Austin Bakersfield Baltimore Baton Rouge Beaumont Berkeley Birmingham Buffalo Charlotte Chesapeake Chicago Cincinnati Colorado Springs Columbia Columbus GA Columbus OH Concord Corpus Christi Dallas Denver Des Moines Detroit Durham El Paso Erie Flint Fort Lauderdale Fort Wayne Fort Werth Fremont Fresno Fullerton Garden Grove Garland Gary Glendale APPENDIX Z MINORITY POPULATION 23.59 50.26 30.18 16.68 23.51 16.38 16.22 68.14 14.93 32.76 28.99 56.56 39.37 41.01 36.20 56.77 30.38 33.15 29.42 57.23 35.25 16.23 43.15 37.43 24.25 14.27 52.70 43.10 33.68 10.64 66.59 48.54 67.09 11.46 44.79 25.54 17.74 36.51 24.60 37.30 21.04 21.70 13.54 78.16 25.62 114 Z MINORITY IN WORKFORCE 15.90 64.22 52.46 16.82 18.60 14.04 36.08 63.34 7.63 39.10 30.40 41.93 58.29 39.65 53.38 32.91 9.35 38.06 27.43 31.91 23.61 13.47 42.65 32.59 22.49 11.50 39.12 37.12 36.89 6.80 59.35 48.07 74.16 5.05 21.03 30.00 15.27 42.63 20.00 40.86 18.81 33.60 10.26 86.34 16.02 Z MINORITY PROT. 12.36 ' 54.34 26.69 14.65 11.93 11.66 11.59 52.71'—' 6.33 27.82 17.26 26.64'15~ 33.33 14.00 35.87 16.03““ 10.06 22.38 9.78 28.13 14.10 10.15 24.20 19.37 13.75 12.97 47.05 18.29 20.00 3.39 35.09‘-— 23.05 62.71 4.11 22.78” 15.00 15.52 30.71 16.67 32.51 8.06 13.07 4.71 73.58 11.05 SVCS. CITY Grand Rapids Greensboro Hampton Hartford Hialeah Hollywood FL Honolulu Houston Huntington Beach Huntsville Indianapolis Irving Jersey City Kansas City MO Knoxville Lansing Las Vegas Lexington Little Rock Long Beach Los Angeles Louisville Lubbock Macon Memphis Mesa Miami Milwaukee Minneapolis Modesto Nashville New Haven New Orleans New York City Newport News Norfolk Oakland Oklahoma City Omaha Orlando Oxnard Pasadena TX Paterson Peoria Philadelphia Phoenix Pittsburgh Portsmouth Providence 115 APPENDIX Z MINORITY POPULATION 20.27 34.64 36.85 55.44 76.06 10.02 72.70 47.76 14.96 22.99 23.40 11.04 51.42 31.91 16.21 22.16 23.46 14.98 33.96 32.31 52.16 29.32 28.36 45.34 28.70 11.97 80.61 28.70 13.33 16.87 24.82 41.22 59.70 48.12 34.70 40.12 65.21 21.32 15.65 34.71 57.36 19.51 63.77 19.24 42.93 21.94 25.74 46.68 21.41 Z MINORITY IN WORKFORCE 12.65 26.85 35.41 34.93 40.56 14.02 82.94 50.02 19.42 21.20 30.23 7.49 16.74 43.98 7.62 12.30 21.99 25.46 26.76 29.39 46.82 30.49 44.77 41.42 21.22 18.38 59.83 17.97 5.97 30.43 27.02 29.00 59.65 40.29 35.20 42.71 50.32 27.82 15.01 44.75 53.35 33.52 26.54 14.02 42.28 30.97 20.89 38.70 10.01 Z MINORITY PROT. SVCS. 11.99"“ 17.50 30.82 25.50 31.97 4.55 75.00 25.40” 5.97 5.97 17.15 2.08 9.12 20.73 2.41 11.55“* 7.10 10.82 20.00 14.55 36.29"' 14.94 8.10 26.57 13.17 10.67 64.29 13.17 9.15 21.79 17.41 18.75 28.74 24.65"" 25.16 11.14 62.12 12.88“”’ 8.58 64.93 24.11 2.81 14.70 8.10 24.01 16.34——~ 12.35 14.41 9.35 CITY Pueblo Raleigh Reno Richmond Riverside Roanoke Rochester Rockford Sacramento San Antonio San Bernadino San Diego San Francisco San Jose Santa Ana Savannah Seattle Shreveport Southbend Springfield MA Stamford Stockton St. Louis St . Pan]. St. Petersburg Sunnyvale Syracuse Tacoma Tampa Tempe Toledo Torrance Tucson Virginia Beach Waco Washington D.C. Waterbury Wichita Winston-Salem Youngstown 116 APPENDIX Z MINORITY POPULATION 38.85 38.24 11.52 52.61 26.44 23.11 32.33 16.43 37.72 62.09 42.95 31.25 47.91 36.16 55.51 50.85 21.62 48.36 21.45 26.16 22.17 42.94 47.16 11.43 19.63 25.34 19.54 17.00 37.46 12.84 21.33 20.72 31.08 14.62 33.44 74.27 19.23 17.01 41.35 18.95 Z MINORITY IN WORKFORCE 36.07 36.07 7.49 74.56 25.55 22.46 17.08 12.79 32.71 71.27 39.13 28.97 49.86 27.78 66.99 53.61 21.55 42.02 21.84 20.57 18.90 34.99 55.76 7.12 28.23 18.86 10.59 12.83 37.24 18.30 12.90 16.27 42.15 21.68 27.77 72.73 15.09 8.71 36.58 18.12 Z MINORITY PROT. 24.38 6.67 3.85 27.81 23.75 6.71 9.99 10.00 27.83 49.94 27.79 18.41 30.01 19.81 64.23 30.81 15.35 18.92 10.44 11.51 16.86 28.73 44.51 9.37 18.58 13.31 5.77 8.39 39.50 12.68 14.55 10.93 21.46 6.84 14.23 43.60 5.23 5.81 16.22 18.09 SVCS. BIBLIOGRAPHY BIBLIOGRAPHY Adrian, Charles R. and Press, Charles. Governing_Urban America. 5th ed. New York: McGraw-Hill Book Company, 1977. Banfield, Edward C. and Wilson, James Q. Citnyolitics. New York: Vintage, 1963. Bardach, Eugene. ”On Designing Implementable Programs.” In Current Issues in Public Administration, 2nd ed. Edited by Frederick 8. Lane. New York: St. Martin's Press, 1982. The Implementation Game. Cambridge, MA: The MIT Press, 1978. Beaumont, Enid F. ”A Pivotal Point for the Merit Concept.” Public Administration Review 34 (September/October 1974): 426-430. Benokraitis, Nijole V. and Feagin, Joe R. Affirmative Action and Equal Opportunity: Action, Inaction, Reaction. Boulder, CO: Westview Press, 1978. Berman, Paul and McLaughlin, Milbrey Wallin. Federal Programs Supporting Educational Change. 8 vols. Santa Monica, CA: The Rand Corporation, April 1975. Vol. 4: The Findings in Review. Federal Programs Supporting_Educatlunal Changg. 8 vols. Santa Monica, CA: The Rand Corporation, May 1978. Vol. 8: Implementing and Sustaining Innovations. Caiden, Gerald E. ”Public Personnel Administration in the Doldrums?” Public Personnel Review 32 (January 1971): 30-35. Cayer, N. Joseph and Sigelman, Lee. ”Minorities and Women in State and Local Government: 1973-1975." Public Administration Review 48 (September/ October 1980): 443-450. Couturier, Jean J. ”Civil Rights in Civil Service-4The Winds of Change.” Public Administration Review 34 (May/June 1974): 243-245. Dahl, Robert A. Who Governs? New Haven: Yale University Press, 1961. Danielson, William F. ”Affirmative Action in the Public Sector: Where Do We Stand Today?” Public Administration Review 34 (May/June 1984): 241-242. Derthick, Martha. New Towns In-Town. Washington, D.C.: The Urban Institute, 1972. 118 119 Dorn, Edwin. Rules and Racial Equality. New Haven and London: Yale University Press, 1979. Downs, Anthony. Inside Bureaucracy. Boston: Little, Brown and Company, 1967. Eisinger, Peter K. Black Employment in City Government, 1973-1980. Washington D.C. in Joint Center for Political Studies, 1983. ' ”Black Employment in Municipal Jobs: The Impact of Black Political Power.” American Political Science Review 76 (June 1982): 380-390). ”The Economic Conditions of Black Employment in Municipal Bureaucracies.“ American Journal of Political Science 26 (November 1982): 774-771. Hall, Grace and Saltzstein, Alan. ”Equal Employment Opportunity for Minorities in Municipal Government.” Social Science Quarterly 57 (March 1977): 864-871. Henkel, Ramon E. Tests of Significance. Beverly Hills: Sage Publications, 1976. Hill, Herbert. ”The Equal Employment Opportunity Acts of 1964 and 1972: A Critical Analysis of the Legislative History and Administration of the Law.” Industrial Relations Law Journal 2 (Spring 1977): 1-96. Hopkins, Anne H. "Perceptions of Employment Discrimination in the Public Sector.” Public Administration Review 46 (March/April 1980): 131-137. Judd, Dennis R. The Politics of American Cities. Boston: Little, Brown and Company, 1979. ’ Kranz, Henry. ”Are Merit and Equity Compatible?” Public Administration Review 34 (September/October 1974). Letts, Richard. Director of Human Relations and Affirmative Action officer, City of Lansing, Interview, 22 September 1983. Levine, Charles H. "Beyond the Sound and Fury of Quotas and Targets." Public Administration Review 34 (May/June 1974): 240-241. Levine, Charles H., Rubin, Irene S. and Wolohojian, George G. ”Resource Scarcity and the Reform Model: The Management of Retrenchment in Cincinnati and Oakland.” Public Administration Review 41 (November/ December 1981): 619-628. Lineberry Robert L. and Fowler, Edmund P. ”Reformism and Public Policy in American Cities.” In Classics of Urban Politics and Administration. Edited by William J. Murin. Oak Park, IL: Moore Publishing Company, Inc., 1982. 120 Lovell, Catherine. ”Three Key Issues in Affirmative Action." Public Administration Review 24 (May/June 1974): 235-237. Lowi, Theodore J. At the Pleasure of the Mayor. New York: The Free Press of Glencloe, 1964. ”Machine Politics-Old and New.” In Bureaucratic Power in National Politics. 3rd ed. Edited by Francis E. Rourke. Boston: Little Brown and Company, 1978. March, James G. and Johan P. Olsen. ”Organizational Learning and the Ambiguity of the Past.” In Ambiguity and Choice in Organizations. edited by James G. March and Johan P. Olsen. Bergen: Universietsforlaget, 1976. Marlin, John Teppen. ”City Affirmative Action Efforts.” Public Administration Review 37 (September/October 1977): 508-511. Mazmanian, Daniel A. and Sabatier, Paul A. Implementation and Public Policy. Glenview, IL: Scott, Foresman and Company, 1983. Miller, Mary Lenn. ”Implementation of Federal Policy at the Local Level: A Study of Affirmative Action in Municipal Governments.” Paper presented at the 1978 Annual Meeting of the Southwestern Political Science Association, Houston, TX, 12-15 April 1978. Milward, H. Bunton and Swanson, Cheryl. ”The Impact of Executive Orders and Administrative Regulations.” Policy Studies Journal 7 (Winter 1978): 201-207. Moore, Wilbert E. (in collaboration with Gerald W. Rosenblum). The Professions: Roles and Rules. New York: Russell Sage Foundation, 1970. Mosher, Frederick C. Democracy and the Public Service. Second Edition. New York: Oxford University Press, 1982. The Municipal Yearbook 1978. Washington, D.C.: International City Management Association, 1978. The Municipul Yearbook 1982. Washington, D.C.: International City Management Association, 1982. Nakamura, Robert T. and Smallwood, Frank. The Politics of Policy Implementation. New York: St. Martin's Press, 1980. Nigro, Felix A. and Nigro, Lloyd G. The New Public Personnel Administration. Itasca IL: F.E. Peacock Publishers, Inc., 1976. Nigro, Lloyd G. ”A Minor Symposium: Affirmative Action in Public Employment, Introduction.” Public Administration Review 34 (May/June 1974): 234-235 . Palumbo, Dennis J. and Styskal, Richard A. ”Professionalism and Receptivity to Change.” American Journal of Political Science 18 (May 1974): 385-394. 121 Peterson, Paul E. City Limits. Chicago: The University of Chicago Press, 1981. Pressman, Jeffrey L. Federal Programs and City Politics: The Dynamics of the Aid Process in Oakland. Berkeley: University of California Press, 1975. ”Federal Programs and Political Development in Cities.” In Improving the Quality of Urban Management. Edited by Willis D. Hawley and David Rogers. Beverly Hills, CA: Sage Publications, 1974. Pressman, Jeffrey L. and Wildavsky, Aaron B. Implementation. Berkeley: University of California Press, 1974. Rich, Wilbur C. The Politics of Urban Personnel Policy: Reformers, Politicans, and Bureaucrats. Port Washington, NY: Kennikat Press, 1982. Rodgers, Harrell R., Jr. ”Fair Employment Laws for Minorities: An Evaluation of Federal Implementation." In Implementation of Civil Rights Policy. Edited by Charles S. Bullock III and Charles M. Lamb. Monterey, CA: Brocks/Cole Publishing Company, 1984. Rodgers, Harrell R., Jr. and Bullock, Charles S. III. Coercion to Compliance. Lexington, MA: Lexington Books, 1976. Rodgers, Harrell R., Jr. and Bullock, Charles S. III. Law and Social Change: Civil Rightp Laws and Their Consequences. New York: McGraw-Hill Book Company, 1972. Rosen, Bernard. ”Affirmative Action Produces Equal Employment Opportunity for All." Public Administration Review 34 (May/June 1974): 237-239. Saltzstein, Grace Hall. "External Pressures, Organizational Constraints, and Change: The Case of Municipal Affirmative Action.” Paper presented at the 1983 Annual Meeting of the American Political Science Association, Chicago, IL, 1-4 September 1983. Sayre, Wallace and Kaufman, Herbert. ”Administrators of the Line Agencies." In State and Urban Politics. Edited by Richard I. Hofferbert and Ira Sharkansky. Boston: Little, Brown and Company, 1971. Scott, W. Richard. Orgenizations. Englewood Cliffs, NJ: Prentice-Hall, Inc. 1981. Shafritz, Jay M. Public Personnel Management: The Heritage of Civil Service Reform. New York: Praeger Publishers, 1975. Sigelman, Lee. ”The Curious Case of Women in State and Local Government." Social Sciences Quarterly 50 (March 1976): 591-604. Stahl, 0. Glenn. Public Personnel Administration, 4th ed. New York: Harper & Brothers, 1956. Public Personnel Administration, 7th ed. New York: Harper and Row Publishers, 1976. 122 Stone, Clarence N.; Whelan, Robert K. and Murin, William J. Urban Policy and Politico_lu a Bureaucratic Age. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1979. Stone, Donald R., Price, Douglas and Stone, Kathrynn. ”City Manager Government.” In Classics of Urban Politics and Administration, Edited by William J. Jurin. Oak Park, IL: Moore Publishing Co., Inc., 1982. Thompson, Frank J. ”Civil Servants and the Deprived: Socio-political and Occupational Explanations of Attitudes Toward Minority Hiring.” American Journal of Political Science 22 (May 1978): 325-347. Personnel Policy in the City: The Polltics of Jobs in Oakland. Berkeley: University of California Press, 1975. Thompson, Frank J. and Browne, Bonnie. "Commitment to the Disadvantaged Among Urban Administrators: The Case of Minority Hiring." Urban Affairs Quarterly 13 (March 1978): 355-377. U.S. Commission on Civil Rights. For All the People...By All the People: A Report on Equal Opportunity in State and Local Government Employment. Washington, D.C., 1969. U.S. Department of Commerce, Bureau of the Census. 1977 Census of Governments. 9 vols. Vol. 3: Public Employment. U.S. Department of Commerce, Bureau of the Census. 1980 Census of Population. Vbl. I: Characteristics of the Population: Number of Inhabitants. U.S. Department of Commerce, Bureau of the Census. County and City Data Book 1977. U.S. Department of Labor, Bureau of Labor Statistics. Geographical Profile of Employment and Unemployment 1982. U.S. Equal Employment Opportunity Commission. Affirmative Action and Equal Employment: A Guidebook for Employers. Washington, D.C.: 1974. Warren, George A. Detroit Civil Service Administrator, 1940-1976. Personal Interview, March 4, 1983. Welch, Susan and Karnig, Albert K. "The Impact of Black Elected Officials on Urban Expenditures and Intergovernmental Revenue.” Paper presented at the 1978 Annual Meeting of the American Political Science Association, New York, NY, 31 August - 3 September 1978. Wilson, James Q. "Innovation in Organization: Notes Toward a Theory.” In Approaches to Organization Desigh. Edited by James Thompson. Pittsburgh: The University of Pittsburgh Press, 1966. Varieties of Police Behavior. Cambridge: Harvard Press, 1978. 123 Wirt, Frederick M. Power in the City: Decision Making_in San Francisco. Berkeley, University of California Press, 1974. Wolfinger, Raymond E. The Politics of Progress. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1974. Wolkinson, Benjamin W. Blacksy_Unions, and the EEOC. Lexington, MA: Lexington Books, 1973. Wurf, Jerry. ”Merit: A Union View.” Public Administration Review 37 (September/October 1974): 431-439. Yin, Robert K. ”Studying the Implementation of Public Programs.” In Studying Implementation: Methodologleal and Administrative Issues. Chatham, NJ: Chatham House Publishers, Inc., 1982.