THE EFFECT 0;: 00mm ACCEPTANCE ; ON TRANSFER OF Rasauacss 1- * ANI) GBALAT‘I‘AINMENT: f w _ A STUDY WINTER-ORGANIZATIONAL _ g ’ ~RELATIONSHIPSV- ' " - Dissertaition for me Degree of Ph’. D._ ‘ , , - MICHIGAN STATE UNIVERSITY: " _' REGINALD KENNETH CARTER ' ‘1974_ 5" fauna-aim. 2.91”“??? mush 1' I1 5‘. " .....',;4 3...? .'2',.‘ .1. z .: “.331 K This is to certify that the thesis entitled ' THE EFFECT OF DOMAIN ACCEPTANCE ON TRANSFER OF RESOURCES AND GOAL ATTAINMENT: A STUDY OF INTERPORGANIZATIONAL RELATIONSHIPS presented by Reginald Kenneth Carter has been accepted towards fulfillment of the requirements for Sociology Ph ' D ' degree in Date /“7'/~9’ 0-7639 in"animus av "5 HUAG & SBNS' - 803V. amum INE I I LIBRARY emom‘ I Q1988”; Mi; 3:; , _'j ”A a} OI p D .r n W ABSTRACT THE EFFECT OF DOMAIN ACCEPTANCE ON TRANSFER OF RESOURCES AND GOAL ATTAINMENT: A STUDY OF INTER-ORGANIZATIONAL RELATIONSHIPS BY Reginald Kenneth Carter Domain acceptance is the approval of organizational domain claims by another organization. It is considered by most interorganizational analysts a requisite for the necessary exchange of resources for adequate goal attainment under conditions of economic scarcity. This important variable has remained unexamined empirically in terms of either its determinants or consequences. This research adepts the three dimensions of domain (i.e., goals, services and client population) outlined by Levine and White (1961) as important determinants of domain acceptance. The perceptions of employees of set member organizations of: (1) the goal attainment level of the focal organization, (2) the ability of the focal organization to deliver its services, and (3) the receptivity of the client population to the domain claims of the focal organization were found to be, individually and in combination, signifi- cantly predictive of domain acceptance. Two other independent Reginald Kenneth Carter variables (i.e., reference group salience of the focal organization and internal consensus) were not found to be significantly related to domain acceptance. The assumed consequences of high domain acceptance were high transfer of resources and, in turn, high goal attainment for the focal organization. The results of this study indicate that high domain acceptance leads to an effective level of resource transfer for maximum goal attain- ment. The effective level was not the highest amount of resources transferred. An excessive amount of resources which were transferred was characteristic of low domain acceptance and was interpreted as an attempt to insure program failure in the focal organization by needlessly increasing the amount of bureaucratic processing of clients. The effective level of resources transferred allowed the focal organization to perform its task and obtain high goal attain- ment. The research was conducted in eight Michigan cities in two large public agencies (i.e., Michigan Employment Security Commission and Department of Social Services) which were legally required to co-ordinate their efforts in a joint training and employment program for welfare recipients called the WIN program. The results of the study indicate a need for inter- organizational researchers and theorists to assess the applicability of "exchange models" as explanatory frameworks Reginald Kenneth Carter since they seldom include conflict as an important inter- organizational reality. The strong ideological emphasis on the positive values of sharing, co-operation, co-ordination and integration reinforce the use of the "exchange models." However, the results of this research, as well as other findings discussed in the report, demonstrate the need for an expanded theoretical framework which is able to incorporate both elements of co-Operation and conflict within the same model. It is to this end that this research has sought to contribute. A typology of inter-organizational relations based on the three dimensions of domain acceptance (i.e., goals, services, client population) was proposed to suggest future areas of research. Illustrations of the more common types were offered in order to demonstrate the usefulness of this typology for predicting transfer rates and goal attainment for the agencies involved. THE EFFECT OF DOMAIN ACCEPTANCE ON TRANSFER OF RESOURCES AND GOAL ATTAINMENT: A STUDY OF INTER-ORGANIZATIONAL RELATIONSHIPS BY Reginald Kenneth Carter A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1974 ACKNOWLEDGMENTS I would like to thank Professor Philip Marcus, my dissertation chairman, for his many helpful suggestions and encouragement throughout this research project. I appreciate, in particular, his extra efforts involved in directing the thesis during my venture several hundred miles away at the University of Wisconsin-Parkside. I am also grateful to the other members of my committee, Professors Harry Perlstadt, Donald Olmsted, Christopher Sower, for their very thorough reading of the proposal and earlier draft plus their thoughtful, insightful criticisms which have improved my understanding of the inter-organizational reality I have sought to explain. Their comments, along with Professor Marcus', have been instrumental in significantly improving the quality of the thesis. I also am thankful to Professor William Faunce who read an earlier draft of the proposal and forwarded his evaluation of it to me while he was on sabbatical in England. Two members of the staff of MESC were particularly help- ful in collecting the data for this study: Mr. Herb Shefline, WIN State Coordinator and Miss Nancy Felder of the Research and Statistics Department. I especially would like to thank my wife, Judy, for her comments, proof reading assistance and patience in helping me finish the thesis. ii TABLE OF CONTENTS Page ABSTRACT ACKNOWLEDGMENTS ................ . ...... . .............. ii LIST OF TABLES ..... . ........... . ......... . ........... v LIST OF FIGURES ..... 0.0.0.000...0.00.00.00.00....0...Vii].- CHAPTER ONE: STATEMENT OF THE PROBLEM AND REVIEW OF LITERATURE.0.0.0....OOOOOOOOOOOOCOOO l I. STATEMENT OF THE PROBLEM..... ...... 1 II. THEORY AND RESEARCHOIOOOOOOOO ...... 3 A. Domain, Domain Consensus, Domain Acceptance.............. S B. Factors Determining the Degree of Domain Acceptance........... 9 1. Set member characteristics. 10 2. Focal organization characteristics............ 13 C. Exchanges and Transfer of Resources.OOOOOOOOOOOOOOOOOOOOO 14 D. Factors Determining the Kind and Amount of Resources TranSferred.OOOOOOOOOOOOOOOOOOO 18 1. Set member characteristics. 18 2. Focal organization CharacteristiCSOOOOOO0....O 24 3. Factors beyond the set..... 25 E. Goal Attainment....... ......... 25 F. Factors Determining Goal AttainmentOOOOOOOOOOOO ......... 27 III. SUMMARY AND CONCLUSIONS ............ 28 CHAPTER TWO: COLLECTION OF DATAOIOOOOOOOOOOOOOOIOOOO 30 I. FOCUS OF STUDY: THE WIN PROGRAM... 30 II. THE SAMPLE.. ................... .... 33 iii TABLE OF CONTENTS (continued) III. THE QUESTIONNAIRE.......... ..... IV. SUMMARY AND CONCLUSION.......... CHAPTER THREE: OPERATIONALIZING THE CONCEPTS AND INDEX CONSTRUCTION..CCCOOCOOOOOCOOOO I. OPERATIONALIZING THE CONCEPTS... A. B. C. D. Domain Acceptance........... Factors Determining Domain Acceptance.................. Transfer of Resources....... Goal Attainment............. II. PROCEDURE FOR THE DEVELOPMENT OF INDEXESOOOOOOOOOICOOOOODOC... III. THE INDEXESOOOICCOOOOOOOOO00.... A. B. C. G. H. Domain Acceptance........... Perceived Goal Attainment of the Focal Organization...... Perceived Confidence in the Focal Organization's Ability to Deliver Its Services..... Perceived Client Receptivity to the Focal Organization's Domain Claims............... Co-operation and Support Offered by the Focal Organization................ Exchange of Ideas, Opinions, Information Between the Focal Organization and the Set Member.................. Salience of the Focal Organization as a Reference Group....................... Internal Consensus... ....... IV. SUMMARY AND CONCLUSION.......... CHAPTER FOUR: DETERMINANTS OF DOMAIN ACCEPTANCE.... CHAPTER FIVE: DOMAIN ACCEPTANCE, TRANSFERS AND GOAL ATTAINMENTOOOOOOOOOOOOOOOOOOOOOOOOIOO CHAPTER SIX: CONCLUSIONS AND FUTURE RESEARCH ...... BIBLIOGRAPHY................................ ...... .. iv Page 41 43 44 44 44 45 49 50 S4 56 56 58 58 68 70 74 77 79 82 86 109 133 146 Table 1. 10. 11. 12. LIST OF TABLES Page Total WIN Team Members and Staff of DSS Functioning in Some Aspect of the WIN Program and Percent Participating in this Study........ 34 Participants in This Study from D58 and WIN Programs in Eight Cities....................... 36 WIN Program Information by City (June, 1970)... 37 Percent and Ranking of Domain Acceptance by 088 county OfficeOOOOOOOOOOOOOOOOOOOOOOOOOIOOOO 57 Mean and Rank of Perceived Goal Attainment of Focal Organization by DSS County Office........ 59 Percent and Rank of First Measure in Index of Confidence in Focal Organization's Ability to Deliver Its Services by DSS County Office...... 60 Means and Rank of Second Measure of Index of Confidence in the Focal Organization's Ability to Deliver Its Services by DSS County Office... 62 Means and Rank of Third Measure in Index of Confidence in the Focal Organization's Ability to Deliver Its Services by 088 County Offices.. 64 Ranks of Two Measures Used to Form Final Index of Confidence in Focal Organization's Ability to Deliver Its Services by 088 County Office... 67 Mean Responses of Four Items Used to Construct Index of Perception of Client's Receptivity to Focal Organization's Domain Claims by DSS County Office........ ........ .................. 69 Item Inter-correlations for Items Comprising the Index of Perception of Client's Receptivity to Focal Organization's Domain C1aims.......... 71 Means and Ranking of Three Items Used to Construct Index of Co-operation and Support Offered by the Focal Organization by DSS County Office................... ....... ............... 73 Table 13. 14. 15. 16. l7. 18. 19. 20. 21. 22. 23. 24. LIST OF TABLES (continued) Item Inter-correlations for Items Comprising the Index of C00peration and Support Offered by the Focal Organization..................... Means and Ranking of Three Items Used to Construct Index of Exchange of Ideas, Opinions and Information Between the Focal Organization and the Set Member by DSS County Office ..... .. Item Inter—correlations for Items Comprising Index of Exchange of Ideas, Opinions and Information Between the Focal Organization and the Set Member............................ Means and Ranking of the Three Items Used to Construct the Index of Salience of Focal Organization as a Reference Group for DSS by DSS County Office............................. Item Inter-correlations for Items Comprising the Index of Salience of Focal Organization as aReference GrouPCOOCOOO00.000.00.000...0.0... Percent and Rank of Internal Consensus by WIN Team Members in MESC Branch Offices........... Ranking of Domain Acceptance and Determinants of Domain Acceptance by City.................. Correlations (r) Between Domain Acceptance and Determinants of Domain Acceptance......... Ranking of Perception of Goal Attainment, Ability to Deliver Its Services, Receptivity of Clients to the Focal Organization's Domain Claims by DSS County Office................... Comparison of Four Expected Equations for Testing Accuracy of Causal Model With Four Actual Equations Drawn From Our Research Data. Revised Comparison of Four Expected Equations for Testing Accuracy of Causal Model With Four Actual Equations Drawn from Our Research Data.0....OOOOOOOOOOCOOOOO0.000000000000000... Mean and Rank of Salience of Reference Groups for DSS Employees in Eight DSS County Offices. vi Page 75 76 78 80 81 83 85 87 91 99 103 105 Table 25. 26. 27. 28. 29. 30. 31. 32. LIST OF TABLES (continued) Page Referrals Per Team and Ranking of Referrals Per Team by WIN Program...................... 110 WIN Program Goal Attainment From Start of Program to June, 1970. ........ ..... .......... 112 Ranking of Goal Attainment by WIN Program.... 114 Ranking of Domain Acceptance by DSS County Offices; Referrals Per Team and Goal Attainment by Corresponding WIN Program...... 115 Effect of Domain Acceptance and Referrals Per Team on Goal Attainment of WIN Program....... 119 WIN Enrollee Characteristics by WIN Program From Start of Program to June, 1970.......... 121 Partial Correlation Coefficients Between Referrals Per Team and Goal Attainment, Controlling for Sex, Age, Race and Education of Client.................................... 123 Comparison of Four Expected Equations for Testing Accuracy of Causal Model With Four Actual Equations Drawn from Our Research Data ....... . ........ ... ........ ......... ..... 128 vii LIST OF FIGURES Figure 1. Litwak and Rothman's Hypothesized Conditions for the Most Efficient Level of Formality of EXChange LinkageOOOOOO......OOOOOOOOOOOOOOIOOO Hypothesized Relationships Between the Factors Determining Domain Acceptance, Transfer of Resources and Goal Attainment................. A Brief Description of WIN Team Member's Main Duties and Primary Inter-organizational contaCtSOOOOOOOO0.0.0.0.......00.00.000.000... Correlations and Partial Correlations Between Co-operation and Support, Exchange of Ideas, Opinions and Information, Reference Group Salience and Domain Acceptance................ Revision of Figure 4 According to Blalock's MOdel.................COOOOCO......OOIOOOIOOO. Causal Model of Determinants of Domain Acceptance and Consequences of Domain Acceptance for Referral Rate and Goal Attainment.................................... Revised Causal Model of Determinants of Domain Acceptance and Consequences of Domain Acceptance for Referral Rate and Goal Attainment.......................... ..... Typology of Inter-organizational Relations According to Dimensions of Domain Acceptance and the Consequences for Transfer of Resources and Goal Attainment..... ..... ................. viii Page 21 29 39 98 102 125 127 146 CHAPTER ONE STATEMENT OF THE PROBLEM AND REVIEW OF LITERATURE I. STATEMENT OE THE PROBLEM Emile Durkheim, in his book on The Division of Labor in Society (1933) attempted to explain how various functions within a society were assigned, maintained and co-ordinated. With the tremendous growth in the size and structure of the labor force, the unit of analysis has shifted from Durkheim's occupational groups to organizations. Consequently, many of the issues which Durkheim addressed are being re-examined in the theory and research on interorganizational relations. Two of the important issues which have emerged regarding the division of labor among organizations are: (1) how is one organization assigned a specific task, and (2) how is agreement reached among the remaining organizations regarding the legitimacy of that assignment. This research investigates this second issue by exploring the assignment of the Work Incentive Program (WIN) to the Michigan Employment Security Commission (MESC) and the consequences of degrees of accept- ance of that assignment by the Department of Social Services (DSS). A frequent practice among state and federal governments is to ngassign specific tasks to different agencies with little or no reflection on the negative consequences of such a process. By transferring the responsibility of a task to various organizations at different or at the same time government assures the continued lack of agreement by other organizations regarding the legitimacy of that assignment to a particular agency. With low agreement on legitimacy there is often an accompanying low degree of voluntary exchanges of resources between agencies. Without such exchanges an agency may habitually fail to accomplish its task. Moreover, the entire division of labor among organizations is never allowed to stabilize over a long period of time and thus is not given the Opportunity to become effective. This research examines this process of Egressigning the task of training welfare recipients for jobs (Work Incentive Program) from the Department of Social Services to the Michigan Employment Security Commission. The legitimacy of an organization's domain claims must be approved by the external environment. The acceptance among organizations of another organization's domain claims is called domain consensus. It is conceived by most organizational writers as a requisite for resource exchange and survival under economic conditions of partial inter- dependence. Unfortunately, a lack of systematic research about domain consensus leaves an important gap in our knowledge of organizational relationships. Domain acceptance is part of domain consensus as it limits the approval of an organization's domain claims to one other agency. Domain acceptance does not imply mutual acceptance of domain claims between two or more organizations as is suggested by domain consensus. This research treats domain acceptance as the main independent variable and examines its relationship to the transfer of resources and the focal organization's goal attainment. The major prepositions of this study can be simply reduced to two: Pl: the higher the domain acceptance the higher the transfer of resources to the focal organization; P2: the higher the transfer of resources the higher the goal attainment of the focal organization. There will also be hypotheses proposed to account for variables determining the amount of domain acceptance, transfers and goal attainment. II. THEORY AND RESEARCH The absence of systematic empirical research has not prevented speculation on inter-organizational relations (White and Vlasak, 1970). Several conceptual frameworks, borrowed from different branches of the social sciences, have been used as models to study inter-organizational relations (Marrett, 1971). These include: (1) open-system models (Parsons, 1960; Katz and Kahn, 1966); (2) co-ordination or co-operation models (Litwak and Rothman, 1962): (3) exchange models (Levine and White, 1961); (4) organization-set models (Evan, 1966); and, (5) conflict models (Miller, 1958; Maniha and Perrow, 1965: Hollister, 1970). Two Of these models, Open systems and organization-set models, are particularly relevant to our research. The organization-set models are significant to our research because of our use Of the key concepts Of "focal” and "set" organization throughout the study to indicate specific agencies. The terms organization and agency will be used interchangeably throughout the remainder Of this report. Evan (1966) borrowed the concepts of focal and set from Robert Merton's analysis Of role-set to indicate the interdependence between organizations. Agencies, like roles, cannot be fully understood in isolation but only in terms Of their relationships with other agencies upon which they depend for necessary resources and exchanges in attaining their goals. The focal organization is the focus Of the study. The set organizations are those agencies with which the focal organization interacts to obtain its goals. For example, a public hospital could be a focal organization in a study of health set. Some Of the set members would include: (1) medical supply companies; (2) ambulance services; (3) private medical clinics; (4) health insurance companies; (5) Red Cross; and (6) local churches. The Open system approach is also important because it emphasizes the imperative nature of Obtaining and maintaining organizational legitimacy or the acceptance Of its goals and means by the external environment (Parsons, 1960: 63-64; Perrow, 1970: 92-132). An organization's goals are considered legitimate when they are recognized as useful by the larger social system. Organizational legitimacy, equivalent in inter-organizational literature to supervisory legitimacy in intra-organizational analysis, refers to the acceptance of the right Of the organization to establish procedures to Obtain specific goals. The Open system's framework is seldom clear as to who are the evaluators of legitimacy. Evaluators may include: (1) members Of the society (White, 1968); (2) representatives Of the society (Selznick, 1949): (3) all the organizations that the focal organization interacts with now or in the future (Thompson, 1967:28); and, (4) only the set member organizations (Dill, 1958; Evan, 1966; Duncan, 1972). The primary audience Of evaluation which has been studied by previous inter-organizational investigators is the fourth audience mentioned above. This same audience is the focus Of this research and is, thus, gng_of several possible sources Of legitimacy for domain claims. Acceptance by a single set member about the legitimacy of the focal organization's domain claims is called domain acceptance. Domain acceptance is not synonymous with legitimacy but is only one small example Of it since other audiences could also accept or reject the same domain claims. A. Domain, Domain Consensus, Domain Acceptance Levine and White (1961) define the domain of an organization as the specific goals it wishes to pursue and the functions it undertakes in order to implement its goals. In Operational terms, organizational domain in the health field refers to claims that an organization stakes out for itself in terms Of: (1) disease covered: (2) services rendered: and, (3) populations served. Thompson has suggested that this Operational definition could be applied with appropriate modifications (i.e., substituting range Of products for disease covered) for the analysis of all types Of complex organizations (Thompson, 1967:26) Domain consensus between two organizations refers to "the degree to which they agree and accept each other's claims with regard to problems or diseases covered, services Offered and population served" (Levine, White and Paul, 1963: 1191). MOst writers assume that organizations need to have other agencies accept their domain claims in order tO obtain scarce resources from them for goal attainment (Litwak and Hylton, 1962). Consensus is a very misleading term within the inter- organizational research literature because it has had several Operational meanings depending upon the units Of analysis being investigated. Domain consensus may refer to a mutual acceptance of both the focal organization and 322 set member Of each other‘s domain claims. This dyadic usage is the most common way the term is employed (Levine and White, 1961: Hollister, 1970; Hall, 1973). Secondly, there is set domain consensus in which all the set members accept each other's domain claims or the domain claims Of a specific focal organization (Klonglan et a1, 1973). In this second context domain consensus refers to a characteristic Of an entire set rather than the character- istic Of a dyadic relationship. Our study is restricted to a third and smaller unit Of analysis -— the acceptance by gng_set member Of the domain claims Of a specific focal organization. Rather than call this a third meaning of domain consensus, we shall refer to this third usage as domain acceptance. It does not imply the reciprocal approval by the focal organization Of the domain claims of the set number. Domain acceptance is a character- istic Of the set member -— namely, the amount Of approval it has for the domain claims Of the focal organization. It is also important to recognize explicitly the importance Of using the th£§§_0perationa1 dimensions of domain outlined by Levine, White and Paul (1963) for research on the concept Of domain acceptance. This theoretical separation Of domain into three parts not only helps to clarify some Of the confusion within the literature but also suggests possible determinants Of domain acceptance. Several researchers, for example, use the concept Of domain consensus in a very limiting way by only referring to one Of the three dimensions Of domain. Hollister (1970) in his study of police youth bureaus and juvenile courts Operationalized domain consensus to include only agreement about which organization should perform a specific task. Braito, Paulson and Klonglan (1972) also restricted themselves to the ggal dimension Of domain in their study Of 33 federated health agencies. They investigated the organizations' willingness to include into their domain a future health goal of reducing smoking. TO measure domain they asked the tOp administrator the following single question: "If a new inter-organizational program [i.e., agency] designed to reduce smoking was tO be instigated in this state which organizations should be included in it?“ Each top administrator answered this question along a 5-point Likert scale from "definitely should not" to "defi- nitely should" for his own organization (i.e., self-definition of domain) and for each Of the remaining 32 agencies. The amount Of agreement among these top administrators became a measure Of domain consensus. This measure was positively related to: (l) self-definition Of future domain claims; (2) present amount Of activity for attaining domain claims in this area Of reducing smoking (i.e., anti-smoking school program); and, (3) willingness to allocate resources in this same area. These researchers acknowledged that the goal dimension was only one of the three dimensions Of domain consensus outlined by Levine, White and Paul. The more important limitation of their Operationalization of domain consensus stems from the future imaginary nature Of the focal organization. Braito and her colleagues seriously deviated from the original use Of the term as referring to the amount Of acceptance of goals, clients and services presently being claimed by existing organizations. Thus, the findings Of their research have questionable comparability with the results Of Other studies. In summary, the definitions Of domain and domain consensus Offered by Levine and his associates were considered the most inclusive of those available for the purposes of our research. Such definitions were influential in determining our definition of domain acceptance as the approval Of 3 set member Of the domain claims Of the focal organization in the areas of goals, services and client population to be served. B. Factors Determining the Degree gl Domain Acceptance Since no other inter-organizational writer has exclusively limited himself to domain acceptance the factors determining it have been deduced from the research on domain consensus. Thus, this review assumes that ggm§_of the factors that account for the amount Of domain consensus may also be important in influencing the amount of domain acceptance. Although this is a reasonable starting point, we do gg£_assume that the same factors cause bggh domain acceptance and domain consensus. For example, an osteOpathic clinic may accept the domain claims of a public hospital or A.H.A. (i.e., American Hospital Association) in the hope that the hospital may, in turn, publically recognize the professional status Of osteOpath. The factors responsible for the returned acceptance Of the osteOpathic clinic by the hospital or the A.H.A. will probably be very different (i.e., state legislative pressure not to duplicate medical equipment and facilities). The factors influencing the amount Of domain acceptance can be broadly categorized into two areas: (1) the set member 10 characteristics and, (2) the focal organization character- istics. 1. Set member characteristics Several researchers have suggested that the perceptions of the set member employees may influence the acceptance Of the focal organization's goals, services and client pOpulation. Most Of the studies have indicated that low acceptance Of service claims is a consequence Of differences in perceiving and defining the "proper" solution to the problem to be solved. Miller (1958), for example, contends that the causes Of conflict over the implementation of a delinquency prevention program resulted from differences in "Operating philosophies" (i.e., serviceS). These differences in services resulted from the employees' perceptions Of the etiology Of juvenile delinquency and the best solutions to that problem. Miller assumed that the absence Of differences in perceived cause of delinquency would result in lower conflict, higher co-ordination, exchange Of resources and goal attainment. Hollister (1970) found a similar situation between police youth bureaus and juvenile courts. The police youth bureaus promoted punishment Of juveniles whereas the juvenile courts favored counselling and casework to reduce juvenile delinquency. These conflicts over services Offered resulted from "ideological differences" surrounding the nature and causes of juvenile delinquency. 11 Prior perceptions about the competence Of the focal organization's personnel may also be a cause for low acceptance Of service claims as evidenced by unwillingness tO interact with the focal organizations. Haurek and Clark (1967) contend that: "...if the actor's [i.e., set member] evaluation of personnel [i.e., focal organization] and their role performance is negative, he will be less motivated to initiate interaction and more likely tO avoid it consciously." (Haurek and Clark, 1967:48) The combined research of Miller, Hollister and Haurek and Clark indicates the importance g£_prior perceptions, about the nature and causes of the social problem involved and personnel competence in determining domain acceptance. Although these findings have concentrated on the service dimension Of domain we have extended the discussion Of the causes Of domain acceptance to include prior perceptions Of all three dimensions (i.e., goals, services and client population). The following three hypotheses explicitly consider such perceptions as important possible determinants of domain acceptance. H1: the higher the perceived goal attainment of the focal organization the higher the domain acceptance. Hz: the higher the perceived ability Of the focal organization to deliver its services the higher the domain acceptance. the higher the perceived receptivity of the client population to the goals Of the focal organIEation the higher the domain acceptance. 12 A strong normative emphasis exists within the inter-organizational literature on the positive value Of co-Operation and co-ordination (Black and Kase, 1963; Barth, 1963). The reasons for such co-Operation may include: (1) desire for more resources; (2) seeking to Obtain legitimacy from rest Of set (i.e., osteopathic hospitals seek co—Operation from Visiting Nurses Association -— Levine, White and Paul (1973); and (3) expressing their avowed organizational goals (i.e., community planning council wants co-Operation from all community agencies). Regardless Of the reasons for co- operation many interorganizational writers, as well as social policy analysts, assume that co-Operation will lead to higher acceptance Of domain claims, exchanges in resources and eventual goal attainment. The explicit sequence by which this "magical" process develops is seldom elaborated. One causal sequence investigated in this research suggests that a focal organization which is perceived as highly co- operative encourages frequent exchanges Of ideas, information and Opinions with set member organizations. Such frequent interaction, in turn, may lead to a reference group identification, especially as the goals Of the focal organization are incorporated or accepted as legitimate. Three hypotheses, drawn from reference group theory (Hyman and Singer, 1968) and human relations research (Likert, 1967; Perrow, 1972), summarize this causal linkage: 13 H4a: the higher the perceived co-Operation and support Offered by the focal organization the more frequent the exchange of ideas, informa- tion and Opinions between it and the set member. H4b: the higher the exchange Of ideas, Opinions and information between the focal organization and the set member the higher the salience of the focal organization as a reference group for the set member. H4c: the higher the reference group salience Of the focal organization the higher the acceptance Of its domain claims by the set member. 2. Focal organization characteristics Some characteristics Of the focal organization that have been suggested as important in establishing and maintaining domain acceptance include: (1) prestige Of the sponsors (i.e., Board Of Directors - Ellig and Halebsky, 1961; Katz and Kahn, 1966); (2) extensive and effective public relations activities (Thompson and McEwen, 1958; Levine and White, 1961; Perrow, 1961; Thompson, 1967); and, (3) similar previous and successful domain goals (Hillister, 1970; Braito, et a1., 1972). Most inter-organizational investigators have assumed a consistently high internal consensus, that is, acceptance Of the focal organization's domain claims by its own employees. Varying degrees Of acceptance Of domain claims within an organization, however, may affect the amount Of acceptance found among set member organizations for these same claims. High internal consensus probably leads tO a more consistent and 14 convincing positive picture Of the goals Of the agency. Low internal consensus most likely results in confusion for the set member employees who interact with the focal organization's representatives. Such confusion may, in turn, result in low acceptance Of domain claims by other organizations. Which comes first in the causal chain Of events (i.e., domain acceptance causes internal consensus or internal consensus causes domain acceptance) is very difficult to establish without a longitudinal study design. Without any substantial research in this area to guide our Choice, the researcher is in the awkward position Of simply Offering an ”educated guess" between the "chicken or the egg." For our purposes, and based on the argument outlined above, we have summarized our line Of reasoning in the following fifth hypothesis: H5: the higher the degree Of internal consensus the higher the domain acceptance by the set number. Exchanges and Transfer gngesources Levine and White defined inter-organizational exchange as "any voluntary activity between two organizations which has consequences, actual or anticipated, for the realization of their respective goals or Objectives" (Levine and White, 1961: 588). Since our research is only interested in the flow Of resources from the set member to the focal organization, we can define the transfer of resources as voluntary or 15 involuntary activity on behalf Of the set member which has consequences, actual or anticipated, for the realization Of the focal organization's goals or Objectives. Such a defini- tion also deviates from the definition used by Levine and White for exchanges on another issue Of importance. By specifying the voluntary nature of exchange they excluded exchanges that resulted from legal requirements, physical coercion or domination. Unfortunately, they did not justify their arti- ficial limitation. Since it is not Obvious why such a limitation was imposed we have included in our definition Of transfer of resources both voluntary and involuntary activity. Certain transfer Of resources between our organizations in this study were legally mandatory (i.e., referrals). There was, however, considerable variation in the number Of referrals actually made and the amount Of information about the clients supplied by the set member to the focal organization. In short, there were many Opportunities for the members of the set organization tO increase or decrease the quantity and/or quality Of resources transferred within certain legal parameters. This variability in the amount of resource transferred under ”forced co-ordination" is not uncommon. Hollister, for example, in his study Of police youth bureaus and juvenile courts found a similar situation. Where legal requirements were broad and/or ambiguous the Juvenile Court rejected over half Of the referrals sent by the Forest City Youth Bureau on the basis of insufficient evidence. A second illustration 16 intimated that the Juvenile Court did not systematically inform the police youth bureaus Of the disposition Of the Court on referrals on the basis that this was confidential information. Hollister summarized this aspect of selective exchange in the following fashion: "...it appeared that neither information nor cases were exchanged as much as they might have been had there been greater domain consensus among the three organizations." (Hollister, 1970: 11) The units being transferred may include: (1) referrals Of cases, clients or patients; (2) labor services including volunteer, clerical or professional personnel; and (3) funds, equipment and information on cases or technical matters (Levine and White, 1961). Finally, exchanges can be non-reciprocal (i.e., flow is only one way), reciprocal (i.e., symmetrical flow Of resources between both organizations) or joint (i.e., flow from each organization in a set to a third party). Since we are investigating the acceptance Of the focal organization's domain claims by a set member our assessment Of the exchange process is only limited tO a non-reciprocal exchange of resources from the set member to the focal organization. This is a short-run perspective which only measures the transfer Of referrals from the Department Of Social Services (i.e., DSS - the set member) to the Michigan Employment Security Commission's Work Incentive Program (i.e., WIN - the focal organization) which in turn trains and places welfare recipients in jobs. 17 Such a transfer Of clients is probably reciprocal in the long run since the WIN program subsidizes the income Of the welfare recipient while she is participating in the program. This subsidy would, thus, reduce the total grant alotted by DSS to the client. This subsidy by MESC would then be considered a financial savings to the DSS organization. Furthermore, if the client is successfully trained and placed in a job the original welfare recipient may become self- sufficient and eventually terminate as a welfare recipient. This would also be considered as a benefit to the DSS agency. A more realistic expectation is that upon taking a job the client would experience a reduction in the welfare grant, but would probably not become self-sufficient. NOnetheless, this study Of domain acceptance ggly investigates the one-way flow 9: referrals from DSS to MESC-WIN and the consequences Of this transfer on the training and placement process Of these referrals. There is no attempt tO evaluate the goal attainment consequences to the DSS agency (i.e., reduction in welfare recipients or grants) as a result Of the WIN program's level Of goal attainment. Undoubtedly there is some positive relationship between the completion of the two organizations' goals. However, this relationship is 222 investigated in this study. Thus, the whole feedback process which is so important to understand fully the exchange dynamics between two organization is gg£_pursued in this study. Since we are Only concerned with the flow Of resources from one set organization 18 to the focal agency without measuring the reciprocal flow of resources back to the set member the concept Of exchange is not totally applicable. The transfer Of resources is a phrase that is more appropriate to our study and will be used through- out the remainder Of the report to refer to the one-way flow of referrals. D. Factors Determining the Kind and Amount gl Resources Transferred Factors determining the kind and amount Of resources transferred can be categorized into three groups: (1) set member characteristics; (2) focal organization characteristics; and, (3) factors beyond the set. 1. Set member characteristics Levine and White (1961) and Hollister (1970) suggest that domain consensus leads tO high exchanges in resources. Following their lead we suggest that domain acceptance should also lead to high transfers Of resources from the set member to the focal organization. More explicitly, our sixth hypothesis is: H6: the higher the domain acceptance the higher the transfer Of resources from the set member to the focal organization. A second set member characteristic which is considered important for determining the kind Of resources exchanged or transferred is the primary function Of the agency. Levine and White (1961) found that the primary function Of the organizations they studied in the health field (i.e., education, resource, 19 prevention, treatment, rehabilitation) determined the kind Of functional interdependency it had with other set members (i.e., services it could Offer or resources it could receive). This limited the kind of resource an organization could transfer. They found, for example, that treatment organizations ranked highest on the number Of referrals and amount of non-labor resources received whereas educational organizations ranked low on the number Of referrals. A number Of researchers have hinted that complimentary resources are an important requisite for exchange (Barth, 1963; Reid, 1965; Aiken and Hage, 1968). Aiken and Hage maintained that such a choice Of exchange partner (i.e., one which has complimentary resources) ”reduces some Of the problem Of decreased autonomy [as a result Of interdependence in a joint program] because the probability Of conflict is reduced and co-Operation is facilitated in such symbolic arrange- ments." (Aiken and Hage, 1968: 916). Several writers consider the exchange linkages as important determinants or consequences Of the kind and the quality Of units being exchanged (Black and Kase, 1963; Reid, 1965; Aiken and Hage, 1968; Litwak and Rothman, 1970; Klonglan, Rogers and Paulson, 1972; Yep, 1973). Some Of these researchers have developed typologies Of exchange linkages based on the degree of functional interdependence between the organizations 20 involved. Two Of the typologies will be presented tO highlight some Of the analysis in this area. Litwak and Rothman (1970) arrange exchange linkages along a formality continuum from highly informal (i.e., friendship ties, ad hoc luncheon meetings) tO highly formal (i.e., rules, regulations, written agreements). They then hypothesized several conditions that determine the degree Of formality of exchange linkage which would be most effective for the specific organizations being linked. The following page showing Figure 1 summarizes their hypotheses. Litwak and Rothman created a synonym for domain consensus by calling it "symmetrical awareness Of organizational interdependence among set members." Symmetrical meant mutual. Awareness translated into acceptance. Interdependence meant that two or more organizations must take each other into consideration if they are to accomplish their goals. Their original phrase is thus translated into mutual acceptance of each member's goals by the other members in the set gr domain consensus. Consequently, the most efficient exchange linkage between organizations with high domain consensus would be a formal arrangement (i.e., rules, regulations). Litwak and Rothman also described several other dimensions Of exchange linkages such as type Of function served, degree Of autonomy of linkage, authority structure 21 DEGREE OF FORMALITY OF LINKAGE CONDITIONS Informal Formal 1. Degree of Symmetrical Awareness Of Organiza- tional Interdependency Among Set Members (i.e., domain consensus) low high 2. Standardization Of Process of Set Members low high 3. Number Of organizations in Set small large 4. Number Of Exchanges Between Organizations in Set small large 5. Type Of Organizational low degree high degree Structure Of Set Of bureau- Of bureau- Members cratization cratization 1Awareness of interdependency is indicated by the degree Of recognition Of the interdependency (i.e., through the absence Of policy statements about the relationship with another organ- ization - low awareness; or through assigning jobs to deal with another organization - high awareness). Litwak and Rothman (1970) make an unconvincing case for a distinction between symmetrical awareness and asymmetrical awareness (i.e., one organization aware and the other not aware). Although they recognize the inclusion Of organizational awareness in Levine and White's concept Of domain consensus they fail to adequately defend the need to develop a separate concept of organizational awareness of interdependency. Symmetrical awareness Of organ- izational interdependency is translated into domain consensus. Figure 1.--Litwak and Rothman's Hypothesized Conditions for the Most Efficient Level of Formality of Exchange Linkage 22 within the linkage mechanism. Under each condition an appropriate linkage was hypothesized for maximum efficiency. Although these researchers drew from a vast literature to justify their hypotheses there was no systematic attempt to test these proposed relation- ships. Furthermore, it was never clearly demonstrated that such linkages resulted in a more efficient distribution Of resources and/or goal attainment. Both were implied. Klonglan, Rogers and Paulson (1972) have substituted the concept Of strategies for exchange linkages. They have reported an elaborate ordinal continuum Of strategies for resource attainment, and predicted that a more formal strategy led to a more predictable level of resource attainment. These researchers prOposed a Guttman type Of scale Of ”intensity Of co-Operative organizational interaction" which was empirically tested among 18 health and welfare agencies composing health sets in several geographical areas. Their scale revealed the following hierarchy Of 8 items: a. the awareness Of the existence Of the focal organization by the directors Of the set members; b. the acquaintance Of the director Of the focal organization with the directors Of its set members; c. information exchanges between the focal organ- ization and its set members; d. interaction between the director Of the focal organization and the director Of its set members; 23 e. joint programs between the focal organization and its set members; f. resource exchange between the focal organization and its set members (i.e., equipment, personnel); 9. membership Of focal organization's staff on boards, councils, committees Of set members; h. formal written agreements about exchanges Of resources between focal organization and its set members. This scale was unfortunately confusing for our purposes since it did not clearly distinguish between the exchange linkages and the units being exchanged. The exchange linkages can be isolated and ranked from formal to informal tO include Items b, d and h. Any Of these linkages could result in the exchange Of information or resources including participation in joint programs or membership on committees of set members. By combining into a single scale both exchange linkages and units of exchange, these researchers neglected to investigate the amount Of correlation between these twO important sets of variables especially in terms Of the consequences for goal attainment. Other similar scales suffer the same limitation (Black and Kase, 1963; Reid, 1965; Yep, 1973). The only conclusion gleaned from the research Of Klonglan, Rogers and Paulson that is applicable to the exchange linkage concept is that the more formal the linkage or strategy the more predictable the resources attained. In summary, the literature on exchange linkage suggests that the higher the domain consensus the more 24 formal the exchange linkage and the more predictable the units being exchanged. Some Of the consequences Of different exchange linkages included differences in units being exchanged and eventual levels Of goal attainment. The exchange linkages between the two organizations in our research were standardized and formal which indicate that a high degree Of domain consensus may have already existed between these two organizations (Litwak and Rothman, 1970). Two indications of formality were evident: (1) there were numerous regulations regarding the exchange procedure between the two organizations; and, (2) the set member organiza— tion assigned a liaison person to explicitly link the two agencies' exchanges. The very existence of a liaison person represented an organizational measure Of high domain consensus for Litwak and Rothman (1970: 150). Thus, the reader should be aware that this study represents an investigation Of two organizations at the high end Of the domain consensus and domain acceptance continuum. 2. Focal organization characteristics Two focal organization characteristics which are considered determinants Of transfer Of resources are: (1) level Of financial resources outside the set (Levine and White, 1961); and, (2) a long history Of mutually beneficial exchanges in resources (Aiken and Hage, 1968; Aiken and Alford, 1970). 25 3. Factors beygnd the set Factors beyond the set which determine exchanges between organizations include: (1) large social processes such as demographic changes in age composition (Barth, 1963; Clark, 1965; Black and Kase, 1963); and, (2) the law (Hall, 1973). E. Goal Attainment The approach used in this research in studying goal attainment is the more traditional framework which typically defines a goal as "a desired state of affairs which the organization attempts to realize..." (Etzioni, 1964: 6). Writers in this tradition generally equate the terms "goal," "Objective," ”purpose," "mission," "aim," and "task." (Price, 1972) Many factors determine the goal attainment level Of an organization (i.e., efficiency Of technology, resource level, quality Of raw materials, environmental conditions). The main contributing factor investigated in this research is the effect Of transfers Of referrals to the focal organization from the set number on the goal attainment Of the focal organization. Implied in most of the literature on exchanges between organizations is the necessity of such transactions for the goal attainment Of the agencies involved. Klonglan, Yep, Mulford and Dillman (1973) recognized the absence Of any systematic research which assesses the results Of inter-organizational relations on clients or 26 recipients of the services Offered by the organizations involved in the exchange Of resources. Klonglan and his colleagues conducted some research on the outcome Of alcoholic clients as a result Of the quantity Of inter—organizational relations in four different communities. They found: "no clear relationship between the quantity Of inter-organizational relationships and the improved client outcome in these communities. The community with the highest inter-organizational score had the lowest outcome score. However, in the other three communities the rank order of amount Of inter-organizational relations and Client outcome were positively related (i.e., the greater the amount Of inter-organizational relations the greater the client outcome score". (Klonglan, Yep, Mulford, and Dillman, 1973: 10-11) Our research follows the lead of Klonglan and his colleagues by investigating the relationship between inter- organizational relations and client outcome by studying the impact Of the transfer Of resources (i.e., referrals) on the goal attainment Of training and placing welfare recipients in jobs. There are several types Of goals in every organization (Perrow, 1961; Perrow, 1970; 133-174). We shall concentrate on measuring two kinds of goals: (1) long term output goals (i.e., placement of the client in a job); and, (2) short term Operational efficiency (i.e., funds spent per participant). These two kinds Of goals should be positively related. The functional relationship between the units being transferred and the goal attainment measures will be stressed. This is one way Of reducing arbitrary and spurious relationships 27 between these key variables. Moreover, such an approach explicitly assumes that not all goal attainment measures can be directly linked with the units being transferred. F. Factors Determining Goal Attainment Several factors influence the goal attainment Of any organization. Suggested in much Of the inter-organizational research on domain consensus is the positive relationship between the amount of resources Obtained from other organizations and the goal attainment Of the agency receiving such resources. This relationship, however, has not been systematically tested. Our research investigated this relationship. More explicitly, our seventh and final hypothesis is: H7: the higher the transfer of resources from the set member tO the focal organization the higher the goal attainment of the focal organization. Hypotheses six and seven could easily have been "chained" into one two-phase hypothesis: domain acceptance d transfer of resources--’ goal attainment. This is, in fact, the causal chain we are interested in establishing. However, for the convenience Of testing these two causal steps separately we have constructed two distinct hypotheses. Other factors that have been measured in this study which may influence the amount Of goal attainment are some client characteristics (i.e., sex, race, age, education). These factors will be treated as potential intervening variables between the amount Of referrals transferred and the training and placement Of WIN clients. 28 III. SUMMARY AND CONCLUSIONS This first chapter has reviewed the relevant research and theoretical literature pertaining to the determinants and consequences Of domain acceptance. The most important variables have been isolated, defined and evaluated. The following Figure 2 limits these variables to the ones that will be examined in this study. The prOposed relationships between the causes of and consequences Of domain acceptance for transfer rates and goal attainment are important to understand because of the increasing catalytic impact other organizations have in creating internal organizational change (Terryberry, 1968). Moreover, as Aiken and Hage have pointed out, in the future the scarcity Of resources is going to: ...force organizations to enter into more co-Operative activities with other organizations, thus creating greater integration of the organizations in a community structure. The long range consequences Of this process will probably be a heightened co-ordination in communities." (Aiken and Hage, 1968: 928) Before this extensive exchange process becomes institutionalized it would be helpful to clearly understand the most appropriate kind and amount of resources to be exchanged for maximum goal attainment Of the agencies involved. This research study seeks to increase our appreciation of this complex process not only from the perspective of the agencies but also from the perspective of the client. 29 SET MEMBER CHARACTERISTICS l. Perceived Goal Attainment of the Focal Organization (H1) 2. Perceived Ability Of the Focal Organization to Deliver Its Services (H2) 3. Perceived Receptivity TRANSFER of the Client Popula- ACBEIPITANCE” OF ..ATngNMENT tion to the Goals Of. RESOURCES the Focal Organization (H ) (H ) (H3) 6 7 4. Salience Of the Focal Organization as a Reference Group (H4) FOCAL ORGANIZATION CHARACTERISTIC 1. Internal Consensus (H5) . Figure 2.--Hypothesized Relationships Between Factors Determining Domain Acceptance, Transfer of Resources and Goal Attainment. CHAPTER TWO COLLECTION OF DATA I. FOCUS 9; STUDY: THE WIN PROGRAM The focus Of the study was a joint work and training prOgram for clients receiving Aid to Families with Dependent Children called the Work Incentive Program (WIN).1 Although the program was Officially administered by the Michigan Employment Security Commission (MESC) it required the assistance of the Department Of Social Services (DSS) to make referrals and other transfers Of resources (i.e., create a liaison position to facilitate communications with caseworkers and MESC personnel (i.e., WIN team members, interviewers, etc.). 1This research was conducted for the Manpower Adminis- tration Of the Department of Labor and was directed by Professor Philip Marcus under contract number 51-24-69-09. I would like to thank the Department of Labor for funding my research assistantship from 1969 to 1971 under this same contract. During this time my duties included assisting Prof. Marcus in preparing the initial questionnaire and gathering much of the inter-organizational information. A final report of the larger study was prepared by Prof. Marcus (1972). For a more general overview of the WIN program's histor— ical background and a more detailed review Of the program's performance in Michigan see Carter (1970). For a brief description of a typical employment service branch Office see Blau (1963: 1-116). For a longer analysis of interpersonal relations in a social welfare agency see Blau and Scott (1962). 30 31 Historically, DSS had previously administered a very similar program called the Work Experience Program (WEP) from 1964 to 1969. WEP was assessed by the Comptroller General (1969) and evaluated negatively in terms Of effectiveness. This reduction in domain as a result of failure to maintain a certain level of goal attainment supports Hollister's (1970) and Levine and White's (1961) claims that one of the necessary characteristics of domain consensus is the ability to fulfill domain claims. After 1969 the WEP program was transferred from the DSS to the MESC and acquired the new name of WIN. This transfer of domain from one set member to the focal organization was not without consequences as Marcus noted: "In the early stages of the WIN program there was a residue Of resentment between ESC [i.e., Employment Security Commission] at both the state and community level. This resentment stemmed from the fact that a number of HEW Officials [i.e., Department Of Health, Education and Welfare employees], as well as persons connected with the Department of Social Services at the State level, felt that the WIN program had been taken away from them before they had the Opportunity to demonstrate their true competencies. As a result there. was difficulty setting up training programs and Often an uneven set of referrals from DSS to the local ESC office. In some cases, the welfare Offices would dump all available cases on ESC offices; in others there was a lack of referrals." (Marcus, 1972; I-S) The MESC is the focal organization in our study. Its main functions include: (1) distribution of unemployment compensation; and, (2) inform clients of employment Opportunities in the local community. The training and placement of welfare 32 recipients through the WIN program is a relatively new and minor function performed by MESC. The WIN program can best be conceptualized as a relatively autonomous department within the MESC branch Office. The WIN program was usually isolated from the rest of the branch Office Operations. In some cases, it was even physically located in a different building. The branch manager had little authority or control over the Operations of the WIN program which structurally reported to a state co-ordinator who in turn was responsible to the manpower manager. The DSS is ggg_of the set members Of MESC. The main function Of DSS is to distribute social security benefits to welfare recipients or other people who quality for public assistance. We are not concerned with the tglgl Operations of either MESC or DSS. Instead we are only interested in the WIN program's Operations as a specific function of MESC. A complete set Of organizations with which the WIN program must interact for the completion Of its training and placement goals would include: (1) training schools (i.e., beauty salon schools, typing schools); (2) local colleges and high school extension institutions; (3) employers; (4) inter- viewers and labor market analysts of the MESC branch Office; (5) unions; (6) other public agencies (i.e., Department of Vocational Rehabilitation); and, (7) Department of Social Services. Thus, the relationship between MESC's WIN program and the DSS is only one of many inter-organizational linkages 33 that are necessary for attaining its_goals. However, it is a very important relationship since it determines the quality and quantity of welfare clients to be trained and placed in the labor force. This single dyadic relationship between DSS and the MESC's WIN program is the basic unit Of analysis for this study. II . THE SAMPLE Only a small portion Of the Egggl staff Of MESC (i.e., 22% of 307) and DSS (i.e., 39% of 1,043) qualified for inclusion in our study on the basis of actually functioning in some direct Operation Of the WIN program. Of this total Of 479 eligible respondents our research only included 355 who worked most directly with some aspect gf the WIN program. This sample of the total was Egg complete because it did not include those eligible employees who were sick or for some other reason (i.e., caseworker on a house call) were not available to complete the questionnaire when it was administered. Table 1 presents the breakdown of the 479 eligible respondents of which our sample Of 355 was drawn. Approximately 74% Of those civil servants who worked most directly with the WIN program participated in this study. Representatives Of MESC were limited to the WIN team members and representatives of DSS were limited to family and child caseworkers and their supervisors plus WIN liaison personnel. Table 2 presents the actual breakdown of respondents for each DSS county Office and each WIN program in the elght Cities we investigated. The eight cities represented the 34 TABLE 1 TOTAL WIN TEAM MEMBERS AND STAFF OF DSS FUNCTIONING IN SOME ASPECT OF THE WIN PROGRAM AND PERCENT PARTICIPATION IN THIS STUDY PERCENT RESPONDENTS OF GROUP TOTAL IN THIS STUDY TOTAL 1. WIN Team Membersl 68 56 82 2. Staff of DSS: A. Family and Child Caseworkers 334 B. Supervisors of Family and Child Caseworkers2 63 C. WIN Liaison Personnel 14 299 73 TOTAL 479 355 74 1The twelve members Of the WIN teams that were not included in our sample were most Often the clerk-typist or coach. The clerk-typist was not considered, for the purposes Of our research, an important decision-making member of the WIN team. 2We did not have an exact number Of supervisors Of only the family and child caseworkers. Since these caseworkers comprise approximately 63% of the total caseworkers we estimated that 63 Of the 99 total number Of supervisors would approximate the supervisors Of only the family and child case- workers. 35 total number Of cities in Michigan's lower penninsula outside Of Metropolitan Detroit, which had WIN programs at the time Of this research. We decided pg; to include Detroit because we thought that it would be a significantly different and larger urban context than the other smaller cities involved. We also omitted the Upper Peninsula because of its unique geographic, ethnic and economic characteristics (Marcus, 1972). In Table 2 the Q§§ offices are listed by county. The MESC branch Offices are listed by glpy, The jurisdiction Of both Of these sets Of Offices coincide. For example: the second DSS Office is in the county of Genesee and the second MESC branch is in the city of Flint, which is the main city in Genesee County. Royal Oak and Pontiac are both in Oakland County. The WIN programs in larger cities with large welfare rolls were assigned two WIN teams. Table 3 presents some additional information on the WIN programs in the eight cities. Since some cities had two teams we shall refer to WIN programs rather than WIN teams in order to designate WIN Operations in each Of these cities. The entire twelve WIN teams were initiated within a six- month period between September, 1968 to March, 1969. The variation in the amount of funds spent per WIN program was primarily the result Of expenses incurred for training grants. This expenditure generally reflected the percent of clients who were actively participating in some training component. Each WIN team was comprised of six members: (1) team leader; (2) senior counselor; (3) junior counselor; (4) 36 mas 2H3 aza mma some mazmozommmm mamHOHam aaaoa «NH mazmazommmmuzoz mmm mazmozommmm gases mm augmeEOAmmm 2H3 deuce mam muamecoaAmm mma annoy me gases mew gases ml mameOHO .u: an neoomz m OH omflucom A xmo Hmsom as camaxmo a m nuances Hm SeamcH o q aammon .um om sammon .um m m moaamm venue me name a m commxmoz mm Commxmaz m m needs em mommamo N m Bmcwmmm m sunflmmm H Amocooaommomv Asoflumuwcmmuo Hmoomv Amucwosommmmv Ahmoemz uwmv MBHO z mmOHmmo mozamm 0mm: 2 mOHmmo Sazooo mmo 2H m2mmm mHH mm>HAma OH. EHAH: m.onH‘~3lrmty office employees was measured by their mean responses to the following three items: "For each of the following persons or groups how important is it to you that you maintain their good opinion of you and your professional accomplishments?" 78 TABLE 15 ITEM INTER-CORRELATIONS FOR ITEMS COMPRISING INDEX OF EXCHANGE OF IDEAS, OPINIONS AND INFORMATION BETWEEN FOCAL ORGANIZATION AND THE "SET" MEMBER EXCHANGE OF IDEAS, OPINIONS MESC MESC .MESC WIN AND INFORMATION WITH: INTERVIEWERS MANAGER TEAM MEMBERS MESC Interviewers .82 .80 MESC Manager .70 MESC WIN Team Members 79 (a) MESC interviewers (b) manager of the local MESC branch Office (c) WIN team members Table 16 presents the mean responses for these three items. The range of variation of the means among the county Offices for each of the three questions was 1.2 (i.e., from a low of 2.1 to a high Of 3.3), 1.1 (i.e., from a low of 1.8 to a high of 2.9) and .8 (i.e., from a low of 2.5 to a high of 3.3). The sum of the means across the three items also varies considerably from a low of 6.4 to a high of 9.4. This three unit spread indicated a wide distribution along this reference group dimension. The rank of the sum of the means was established by assigning the highest rank (i.e., lst) to the highest sum. This sum indicated a high reference group identification with MESC personnel. Table 17 presents the item inter-correlations for the three questions used to construct the index of salience of MESC for DSS employees. H. Internal Consensus The measure of internal domain consensus was limited to one question which asked the WIN team members if they accepted the domain claims of the WIN program. The amount of internal consensus was indicated by the percent of WIN team anywhere who responded "yes" to the following question: "Do you think the WIN program, in general, that is, on a national scale, is the proper way to handle welfare recipients?" 80 TABLE 16 MEANS AND RANKING OF THREE ITEMS USED To CONSTRUCT INDEX1 OF SALIENCE OF FOCAL ORGANIZATION As A REFERENCE GROUP FOR DSS BY DSS COUNTY OFFICE REFERENCE GROUP IDENTIFICATION WITH: DSS SUM RANK OF COUNTY MESC MESC WIN TEAM OF INDEX OF OFFICE N INTERVIEWERS MANAGER MEMBER MEANS SALIENCE 1 8 3.3 2.9 3.2 9.4 1 2 54 2.4 2.0 2.8 7.2 5 3 39 2.5 2.3 3.2 8.0 3 4 48 3.0 2.5 3.3 8.8 2 5 36 2.6 2.5 2.8 7.9 4 6 31 2.1 2.0 2.8 6.9 6 7 44 2.4 1.9 2.5 6.8 7 8 _32 2.1 1.8 2.5 6.4 8 TOTAL 299 1 The three questions used to establish this index were: "For each of the following persons or groups how important is it that you maintain their good Opinion of you and your professional accomplishments?" Each Option offered was given a numerical value: (5) of very great importance, (4) of great importance, (3) of some importance, (2) of slight importance, (1) of no importance at all. 81 TABLE 17 ITEM INTER-CORRELATIONS FOR ITEMS COMPRISING THE INDEX OF SALIENCE OF FOCAL ORGANIZATION AS REFERENCE GROUP FOR DSS REFERENCE GROUP IDENTIFICATION MESC MESC MESC WIN WITH: INTERVIEWERS MANAGER TEAM MEMBERS MESC Interviewers .92 .76 MESC Manager .83 MESC WIN Team Members 82 Table 18 sets forth the percent of WIN team members who accept the domain claims of the WIN program. There is considerable variation among the WIN teams with the lowest program reporting 67% of its members as approving of the WIN program and the highest programs reporting 100% Of its members accepting the WIN program's goals. One major difficulty with these results is that four of the eight programs are tied for first place with 100% of the WIN members reporting that they accept the program's goals. The way such ties are handled within a ranking procedure is to assign a rank mid- way between the number Of programs that are tied. Since there are four programs tied for first place the mid-way point between 1 and 4 is 2.5. Consequently, these four Offices are assigned the same rank of 2.5. This is an unavoidable limitation of the procedure involved in ranking. Generally, statistical measures are available for correcting correlations for rankings which have ties. Unless there are numerous ties, however, the correctional factor tends to be very small. IV. SUMMARY AND CONCLUSION This first data chapter has operationalized the concepts and described the construction Of indexes. Indexes were formed by grouping those items with high inter-correlations and high face validity. In one case, a set of items were eliminated because of particularly low inter-correlations (i.e., third measure of index of confidence in the WIN program’s ability to deliver its services). Most of the items, however, were retained because of high item inter-correlation and high face validity. 83 TABLE 18 PERCENT AND RANK OF INTERNAL CONSENSUS BY WIN TEAM MEMBERS IN MESC BRANCH OFFICES WIN TEAM MEMBERS IN MESC BRANCH INTERNAL CONSENSUS1 OFFICES N (percent) RANK 1 5 100.00 2.5 2 8 75.00 6 3 9 88.89 5 4 9 100.00 2.5 5 4 100.00 2.5 6 A 5 100.00 2.5 7 10 70.00 7 8 _§ 66.67 8 TOTAL 54 lInternal consensus was indicated by WIN team member's "yes" responses to the following question: "DO you think that the WIN program, in general, that is, on a national level, is the prOper way to handle welfare recipients?" 84 The amount of variation among the DSS Offices or WIN programs in terms Of mean responses was stressed throughout this chapter. The amount of variation on individual questions was often under a full unit along a five point scale. This small amount of variance indicates a serious weakness in the construction of these indexes. However, the high correlation between items retained to construct the index helped to increase the final amount of variance in the index (i.e., sum of means or sum of ranks). This grouping of questions to form a final index was a useful way of organizing small differences in order to uncover a broader pattern of responses. Table 19 presents the results of this Chapter Three by listing the rank order of domain acceptance and the deter- minants of domain acceptance across the eight cities. These rank orders will be utilized in the following two data chapters as the basis for statistically testing the hypotheses suggested in Chapter One. 85 m m.o m.m w m.m m.m m.n c m N w m N e m.c m.o e N m.N m.o n.m o H N N m o m.N m.< c a m.m w m n m m.N m.H N N n.N m.n m.m H w m m m m m.N H H N m o m.q n m n m.n m m N m.N m.H H H w A m.o m H mamoomooo HmaumuoH mucowm moowwm moon» hooowm EOOH>MOM mooowm moz< weHu Hmoom Hmoom Ono mucouomou Hmoow mo muH Hmoowuémmoo< Ono nuHs he oouowmo m we maHmHo um>HHoo mo onus zHazoo soHomsuow uuoooom paw honowm aHMaoo Ou ou madman Ionoum ICH pom :OHooHOAOIOU Hmoom mo aOHuoHomoa Hmoom mo Hmow moOHcHao OOOOHHmm DOOHHO mo muHHHnm oo>Hoouom mmmoH mo EuH>Huooomu oo>Hooumm omcmnoxm oo>Hoouom NHH Mozmo< H mmmzmz HMm VHHU Mm moz<9mmuo< 2Hred' where the same definitions hold for "e", "s” and "d” as mentioned earlier in this paragraph. 96 Given rule number two, the partial correlations should equal to or approach zero. The relationship between co- operation and support E29 salience as a reference group should approach zero or be equal to zero when the intervening variable of exchanges in ideas, opinions and information is controlled for. This becomes translated into rcs.e = 0, where the same definitions for "c", "s" and "e" which were outlined in the last paragraph are maintained. The Kendall partial correlation is only able to consider one intervening variable at one time. Thus, we are unable to estimate the partial correlation between co-Operation and support and domain acceptance when controlling for both the exchange of ideas, opinion and information and reference group salience. By using this statistic we are, thus, limited to the assessment of small three variable linkages within the causal chain. However, if each of these linkages meet the Blalock criteria of an ideal causal model then we should be more confident that the suggested reasoning would apply throughout the causal chain. In summary, if our actual equations compare favorably to the ideal equations suggested by Blalock then we can assume that the causal linkages are reasonably accurate, also assuming that we have included the most important variables in this causal sequence and that our measuring procedure is without serious error. Both of the latter assumptions are tenuous in this exploratory study. Thus, the reader should be cautious about the ability to generalize from the findings 97 of this study to other research settings. We will return to this issue of generalizability later in the report. Figure 4 indicates the correlations among the four variables suggested in hypothesis four. The causal model implied was assessed by comparing the expected relationships among the variables with the actual relationships reported in this study. If our model approximates the relationships outlined by Blalock for a causal model, then we can have some additional confidence in the explanation being offered. . Table 22 compares the four expected equations for testing the accuracy of a causal model with the four actual equations established from our research study in terms of "degree of fit." Degree of fit refers to how closely the actual equations approximate the ideal of expected equations. The degree of fit assessment into the three categories of "low,” ”moderate," and ”high" is an arbitrary judgement by the researchers. There is no consensus or rules of thumb to guide researchers in this area. The reader should thus be aware of this judgemental factor in such an assessment. In comparing the actual relationships in our model with the expected or ideal causal model relationships there are several obvious shortcomings. In equation one in Table 22 the relationship between co-operation and support and exchange in ideas, opinions and information (i.e., two adjacent variables in the causal chain) was r = .71, which is slightly lower than the relationship between co-Operation and support and reference group salience 98 r = .36 r = .18 /—-———. \\ EXCHANGE -~1£> CO-OPERATION'~——{> 0F IDEAS, REFERENCE DOMAIN AND r = .71 OPINIONS r = .86 GROUP r = .29 ACCEPTANCE SUPPORT AND SALIENCE INFORMATION r = .79 PARTIAL CORRELATIONS: (1) rCS e = .40 c = co-operation and support = _ e = exchange of ideas, opinions and (2) red.s '13 information 3 = reference group salience d = domain acceptance Figure 4.—-Correlations and Partial Correlations Between Co-Operation and Support, Exchange of Ideas, Opinions and Information, Reference Group Salience and Domain Acceptance 99 TABLE 22 COMPARISON OF FOUR EXPECTED EQUATIONS FOR TESTING ACCURACY OF CAUSAL MODEL WITH FOUR ACTUAL EQUATIONS DRAWN FROM OUR RESEARCH DATA EQUATIONS DEDUCED FROM BLALOCK'S RULES FOR AN IDEAL CAUSAL MODEL ACTUAL EQUATIONS BASED ON RESULTS ACCURACY OF OUR RESEARCH "Degree of Fit” 1. rce’rcsPrcd .71 .79 .36 low 2. re§}red .86 .29 high 3. res.?0 .40 low (rcs = .79) 4. red.s"'o -.13 moderate (red = .18) CONDITIONS = co-operation and support = exchange of ideas, opinions and information 3 = reference group salience d = domain acceptance 100 (i.e., .79). Reference group salience is one step further along the causal chain and, thus, its correlation to co- operation should ideally have been smaller rather than larger than the correlation between co-operation and exchange of ideas. The conclusion that we drew from this comparison of equation one between the ideal and the actual results was that there was a low degree of fit. Consequently, this finding suggests that the model we have prOposed is not partic- ularly close to the ideal causal model outlined by Blalock. Equation two, on the other hand, does meet the require- ments of the ideal causal model. It is assessed as having a high degree of fit with the ideal. Equations three and four are both partial correlations which should be equal to or approach zero. Both approach zero but do not equal it. Equation three was assessed as having a low degree' of fit with the ideal model's expectation because of the large size of the remaining partial correlation (i.e., r = .40). Any partial above a t .20 was assessed as having a "low” fit with the ideal expectation of zero. Any partial below this level but above the j .10 level was assessed as ”moderate." Finally, any partial between this level and zero as well as including zero was assessed as having a ”high" degree of fit with the ideal partial equations suggested by Blalock. Equation four was assessed as having a "moderate fit" ‘dith the ideal model because it approached zero (i.e., -.13). 101 The third rule outlined by Blalock indicated that the variables used in a causal chain should not measure the same underlying dimension. The fact that there were considerably high correlations between nonadjacent variables indicated that the explanatory variables may not have been distinctively different enough from each other to establish a clear causal chain. The variables appear to overlap each other along a common dimension (i.e., co-operation). In summary, the causal model suggested by our research does not meet very many of the ideal requirements for such a model. More specifically, regarding H4C' the salience Of the focal organization as a normative reference group was 222 found to be significantly related to domain acceptance (i.e., r = .29, p.‘20). This indicated that identification with another organization as a reference group did ngt_necessarily mean that the employees of the set organization accepted the goals of the focal organization. In order to further investigate the inability of our model to approximate the ideal conditions outlined by Blalock we revised Figure 4 by re-arranging the variables into a statistical sequence which satisfies Blalock's first rule. The results are shown in Figure 5. Table 23 indicates the "degree of fit" of the revised model based on Blalock's rules for model construction. This revised model more closely approximates Blalock's requirements. This model clearly removes reference group identification to the outer edges of the causal chain. Such 102 r = .29 r = .18 \ EXCHANGE REFERENCE ————> OF IDEAS, ,___? CO- OPERATION DOMAIN GROUP 1' OPINIONS r r = .36 ACCEPTANCE SALIENCE SUPPORT INFORMATION r = .79 PARTIAL CORRELATIONS: (l) rSC e = .40 s = reference group identification _ _ e = exchange of ideas, opinions (2) red.c '33 and information c = co-Operation and support d = domain acceptance ZFigure S.--Revision of Figure 4 According to Blalock's Model 103 TABLE 23 REVISED COMPARISON OF FOUR EXPECTED EQUATIONS FOR TESTING ACCURACY OF CAUSAL MODEL WITH FOUR ACTUAL EQUATIONS DRAWN FROM OUR RESEARCH DATA EQUATIONS DEDUCED FROM BLALOCK'S ACTUAL EQUATIONS RULES FOR AN IDEAL BASED ON RESULTS ACCURACY CAUSAL MODEL OF OUR RESEARCH "Degree of Fit" 1. rse,rsc’rsd .86 .79 .29 high 2. rec’red .71 .18 high 3. rsc.e-.0 .40 low (rSC = .79) 4. redngO -.33 low CONDITIONS s = reference group salience e = exchange of ideas, Opinions and information c = co-operation and support d = domain acceptance 104 causal remoteness is characterized by the low correlation between reference group identification and domain acceptance (r = .29). Such a low correlation necessitates the serious questioning of the relevance of this concept in explaining the occurrence of domain acceptance. One important factor that may explain the low predictive power of reference group identification with MESC as a determinant of domain acceptance is the relatively low position 23 MESC among all other reference groups reported by DSS employees. The following Table 24 indicates the hierarchy of DSS reference groups in order of their salience or importance as a source of self-concept. The three representatives of MESC (i.e., WIN team members, interviewers and manager of the local office) were ranked in the lg§£_four groups listed in Table 24. This low relative ranking suggested that most DSS employees did not place much importance on the Opinion of MESC personnel regarding their own (i.e., DSS employees) professional activities and self-concept. One unexpected result of assessing the causal model implied in the fourth hypothesis on the relationship between reference group salience and domain acceptance was that 2222 of the variables were strongly related to domain acceptance. As reported earlier reference group salience was not significantly related to domain acceptance (i.e., r = .29, p.‘: .20). Moreover, neither co-operation and support nor exchange in ideas, Opinions and information were significantly H 105 TABLE 24 MEAN AND RANK OF SALIENCE OF REFERENCE GROUP FOR DSS EMPLOYEES IN EIGHT COUNTY OFFICES (N = 299) REFERENCE GROUP SALIENCEl RANK 1. DSS Supervisors of Social Workers 4.1 1 2. DSS Unit Supervisors 3.9 2.5 3. DSS Local Officials 3.9 . 4. WIN Applicants 3.8 4 5. DSS Social Workers 3.7 5 6. DSS Liaison for WIN 3.3 6 7. Training Agencies 3.1 7 8. DSS State Officials 3.0 8 9. WIN Team Members 2.8 9.5 O. Employers in Local Community 2.8 9.5 11. MESC Interviewers 2.4 11 12. MESC Manager 2.1 12 1 This data was drawn from DSS employee responses to the following question: "For each of the followinggroups or persons how important is it that you maintain their good opinion of you and your professional accomplishments?" The respondents were given five options to choose from. Each option was assigned a numerical value: (5) of very great importance, (4) of great importance, (3) of some importance, (2) of slight importance, (1) of no importance at all. 106 related to domain acceptance (i.e., r = .36, p.(:.l4; r = 18, p.{‘.30). This finding seriously contradicts the strong normative emphasis within inter-organizational relations literature on both high co-operation and high exchange of resources between organizations. We shall return to this issue in the final chapter of this report. The last determinant of domain acceptance examined in this research is a focal organization characteristic - internal consensus. Internal consensus is the degree to which members of the focal organization accept the goals of their own organization. Table 20 showed the correlation between internal consensus and domain acceptance as only moderate (i.e., r =.25, p.(.27). The reasoning that led to this unconfirmed hypothesis must now be reconsidered. The amount of acceptance by the focal organization's employees was assumed to result in a more consistent and positive image of the WIN program which was presented to the set members. Such a positive image was considered important in establishing the acceptance of the focal organization's goals. However, even when there was 100% reported acceptance in the focal organization there was no guarantee of acceptance of the WIN program by the set member organization. This is an important finding because it suggests that the locus of domain acceptance may not be in the attitudes of the employees of the focal organization and simply transferred to the employees of the set organization via good public 107 relations. Instead, the results of this study indicate that the locus of domain acceptance lies in the selective per- ceptions of the personnel of the set organization surrounding the key components of domain (i.e., goals, services, clients). In summary, this chapter has presented the testing of the first five hypotheses which predicted the determinants of domain acceptance. Three of the hypotheses were confirmed at the .10 level of significance for this exploratory study. These hypotheses suggested that prior perceptions regarding the focal organization's goal attainment, ability to deliver its services and receptivity of the client population to its domain claims were highly predictive of domain acceptance. These findings indicated the importance of such perceptions for establishing domain acceptance. Of the three perceptions regarding the domain acceptance of a focal organization the two most important variables were perceived ability to deliver services and perceived client receptivity to the focal organization's domain claims. These dimensions, in particular, should receive more attention in future research. The final two hypotheses were not confirmed. The salience of the MESC organization as a reference group for DSS employees was not significantly related to domain acceptance (i.e., r = .29, p.(.20). The relatively low salience of MESC personnel among the many reported reference groups was suggested as an important reason for the low predictive performance of this variable. Reference group 108 salience as an important social psychological variable has not been a particularly consistent predictor of behavior within the occupational and organizational context. The most widely used example of reference group identification within this literature is the local-cosmopolitan typology which has only been able to link reference group choice with such behavior as adherence to organizational rules (Gouldner, 1957; Blau and Scott, 1962). Finally, internal consensus, the only focal organiza- tional trait, was also not significantly related to domain acceptance (i.e., r = .25, p.(.27). It appears that the perceptions of focal organizational employees are quite independent of the perceptions of set organization employees regarding the acceptance of the domain claims made by the focal organization. CHAPTER FIVE DOMAIN ACCEPTANCE, TRANSFERS AND GOAL ATTAINMENT This third data chapter treats domain acceptance as an independent variable which determines the amount of resources transferred from a set member to the focal organiza- tion. This transfer, in turn, is assumed to increase goal attainment of the focal organization. This chapter in- vestigates the last two hypotheses in our study: H6: the higher the domain acceptance, the higher the transfer of resources from the set member to the focal organization the higher the transfer of resources the higher the goal attainment of the focal organization In order to test these two hypotheses it was first necessary to arrange the data into rank order across the eight cities. Domain acceptance was already ranked as shown earlier in Table 19. The units being transferred from DSS to the MESC WIN program were referrals. The list of referrals of WIN clients per WIN team plus the subsequent ranking across the eight WIN programs are contained in Table 25. Table 25 shows four Of the WIN programs as having two WIN teams and four with only one WIN team. The referrals were ranked according to referrals per team in order to standardize 109 110 TABLE 25 REFERRALS PER TEAM AND RANKING OF REFERRALS PER TEAM BY WIN PROGRAM WIN PROGRAM REFERRALS PER WIN TEAM1 RANK 1 1,037 l 22 977 2 32 488 8 42 708 6 5 739 s 6 877 3 72 530 7 8 782 4 1 This data was drawn from Felder (1970). 2 These WIN programs have two WIN teams. 111 the effect of different numbers of teams in the eight WIN programs. Finally, the ranking of goal attainment was established by ranking the three measures of program efficiency outlined and Operationalized in the last part of Chapter Three. The raw data used to construct these rankings is presented in the following Table 26. The raw data itself was helpful in understanding some aspects of the WIN program. First, there was a large number of referrals made to the WIN programs. Only a portion of these were enrolled in the program and eventually participated in some training component of the WIN program. Moreover, a very small portion of these participants were ever placed in permanent employment positions. Carter (1970) estimated that only 4% to 8% of all referrals to the WIN program in Michigan ever get placed in the labor force as a direct result of their participation in this manpower program. In almost all WIN programs the number of reported participants in each WIN program exceeds the total enrollees. This was the result of counting the same enrollee each time she enters a new training component. For example, if a client was placed into a training course she was counted as a participant. After completing the training course she then entered an on-the-job training period. She would be counted a second time as a participant. The amount of funds which are specified in Table 26 were funds actually spent by each WIN program. In most programs, 112 .mEmmu 2H3 o3» m>mn mEmHmoum cwz mmmna . thmflv HOUHOW EOHH C3MHU @03 0».va mflSBM oo.mma.mvaw am am Han m¢m th m oo.hmm.momw m.m~ be mmm 0mm coo.H Nb oo.mvm.ammm mm mm mov mum hum o oo.amm.haaw ma ma mom mum mmb m oo.hmm.vmam m.Nm mm mam mmh havta Nv oo.vhm.mnmm m.v¢ mm omm.H mmh ohm mm oo.amm.vmmw m.om Hv Huh 5mm vmm.H Nm oo.mmo.mh w mm mm vmm mmm hmo.H a 82mmm mQZDh dea mmm mBzmSm0¢Am m924mHUHBm4m mmmqqomzm qummmmmm Edmwomm mBzmSmU¢Am 2H3 onma H .mZDh OB Edmwomm m0 Bmflfim 20mm BZHSZHonm om manna mHnu OH mouoaHumo any .ON OHnwa ca HOHHHOO vouuonou saga manna manu OH OOOHHOHOO mo Hanson Howuma mauanHm o oouuocou oowusom poo sowoxnsz .Aonmav Hooaom Bonn gamut mums some omega H Hmo.¢ H4969 o N.GH N N.sH N m.HH s G.oN mmmwu. N N N.mH m H.0H G o.mm H o.am moo N H N.NN N «.4H N m.oN N m.HG mNN G N N.NN H a.om H G.mN o N.GN NNN m e N.oN s m.HN a m.as N m.Nm mNN a m a.oN N N.NN m N.mq m m.mN GNN m m o.qH N G.NN N H.0o N o.mm NNN N N «.mN o N.GH N N.om N m.ao mmN H Mz PERCEIVED ——-? DOMAIN —? REFERRALS = .61 DOMAIN r = .5 ACCEPTANCE r - -.36 (p. <.02) CLAIMS (p. <.03) (p. <.14) Figure 7.--Revised Causal Model of Determinants of Domain Acceptance and Consequences of Domain Acceptance for Referral Rates and Goal Attainment (Kendall Coefficients) 128 TABLE 32 COMPARISON OF FOUR EXPECTED EQUATIONS FOR TESTING ACCURACY OF CAUSAL MODEL WITH FOUR ACTUAL EQUATIONS DRAWN FROM OUR RESEARCH DATA EQUATION'S DEDUCED FROM ACTUAL EQUATIONS BLALOCK'S RULES FOR AN BASED ON RESULTS ACCURACY IDEAL CAUSAL MODEL OF OUR RESEARCH "Degree of Fit" 1. rgp, r95, rgr .61 .54 -.36 high 2. r r .57 -.36 hi h pd, Pr . g 3. r an. 0 .29 low gd-P (rgd - .54) 4. rpLa-OO -.16 moderate CONDITIONS g = goal attainment perceived domain claims "U ll domain acceptance H II referrals per team 129 Equations one and two both met the Blalock require- ments of decreasing magnitude of correlation as the number of intervening steps increase from one causal variable to the next in the causal chain. The partial correlation in equation three was reduced from the original (i.e., from .54 to .29), but remained a considerable distance from zero. The comparison of the actual results with the ideal expected results led to the conclusion that our model only had a "low" degree of fit. Finally, the partial correlation in the fourth equation also showed a reduction in the original relationship (i.e., from -.36 to -.16), but was closer to zero than the partial in the third equation. The actual results of the fourth equation were subsequently assessed as an indication that our model had a "moderate" degree of fit with the ideal causal model requirements outlined by Blalock. In summary, our results indicate that the proposed model does reasonably approximate the ideal requirements for a causal model. However, it does have some important weaknesses. The strength of the relationships between adjacent variables and other variables farther along the causal link is often very similar or only slightly reduced. For example, the correlation between goal attainment and perceived domain claims is .61 whereas the correlation between goal attainment and domain claims is .54. This small difference could be accounted for by chance of measurement error. This kind of weakness should be kept in mind by the reader in assessing the model, especially in terms of generalizations 130 to other dyadic relationships between different organizations than those tested in this report. This revision in the causal chain places more importance on the actual goal attainment levels of the focal organization than our originally proposed sequence. The statistical reality of a high correlation between actual goal attainment level and the perceived domain claims indicates that it is obvious to the set member organization when the focal organization is either successful or unsuccessful in attaining its domain goals. Our revised model suggests that successful completion of domain goals appears to lead to positive appraisal of the goals, competency of personnel and receptivity of client population to the domain claims of the focal organization. This, in turn, leads to high acceptance of these goals by the set member which then refers an apprOpriate level Of clients to the focal organization to maintain high goal attainment. On the other hand, unsuccessful completion of domain goals appears to result in a negative appraisal of the goals, competency of personnel and receptivity of clients to the domain claims of the focal organization. This, in turn, results in low domain acceptance by the set member which then refers an excessive level of referrals to the focal organization to maintain poor program performance. The final difficulty with the simple causal chain analysis which uses correlation coefficients and partial coefficients is that the causal direction is never statistically 131 proven. When a researcher posits that Xi. Y-OZ and presents correlation coefficients to support this claim there are a number Of possible interpretations: (1) X-OYhoZ; (2) Xt-Yo-Z; (3) X-DYo—Z. Since only one of these alternative explanations fits our model (i.e., 1) the only way we can establish that it is preferable to the other two is on the strength of the logical and theoretical argument plus the supporting statistical information. In summary, this third data chapter presented the results of the two final hypotheses regarding the consequences of domain acceptance for referral rate and goal attainment. The raw data for the rankings across the eight WIN programs were presented and the results of the correlations were analyzed. The findings indicated that high domain acceptance led to low referral rates which resulted in high goal attainment. This was the Opposite of the consistently positive relationships suggested by the hypotheses. Such contradictory findings necessitated a revision in our earlier thinking. The new interpretation, based on the results of our study, suggested that the set member employees who reported high domain acceptance also recognized the importance of 22 appropriate level pf referrals for continued program success. This apprOpriate level was not so high or excessive as to be a hinderance to goal attainment, but was high enough to allow full utilization of the potential assistance that the WIN program could offer its clients. 132 The negative relationship between the number of referrals and goal attainment was also further investigated by controlling for such client characteristics as sex, age, race and education. None of the partial correlations indicated any significant impact of these intervening variables on the main relationship between number of referrals and goal attainment. 1:1 Finally, a causal model of the important variables presented in this study was presented, revised and assessed according to rules deduced from Blalock's recommendations for the requirements of an ideal causal model. CHAPTER SIX CONCLUSIONS AND FUTURE RESEARCH NEEDS Many Open system theorists (Parsons, 1960; Katz and Kahn, 1966) and researchers in the area of inter-organizational relations consider organizational goal legitimation by the external environment as an important determinant of intro- organizational processes (Thompson and McEwen, 1958; Dill, 1958; Simpson and Gulley, 1962; Aiken and Hage, 1968; Terryberry, 1968; Yuchtman and Seashore, 1967; Thompson, 1962; Lawrence and Lorsch, 1967) and necessary for organiza- tional survival (Perrow, 1961). Domain consensus, the acceptance of the goals of an organization by the other organizations it must interact with for successful goal completion, has been assumed to be important for obtaining the necessary exchanges in resources from these set members. Only Braito (1972) has attempted to Operationalize domain consensus and to test systematically some of the suggested consequences of this important organizational characteristic. This research has investigated domain acceptance which was defined as the acceptance by 922 set member of the domain claims of the focal organization. Domain acceptance was assumed to be an important component of the concept of domain consensus o 133 134 Our research operationalized the concept of domain acceptance and tested its importance in determining a one way transfer of resources from the set member to the focal organization and the consequences of this transfer for the goal attainment of the focal organization. Domain acceptance was operationalized as a continuous variable rather than simply treating it as an asserted requisite for both interorganizational relations and organizational survival. This research study assessed the degrees of domain acceptance in terms of its determinants and consequences. Three set member characteristics were isolated as determinants of domain acceptance: (1) perceived goal attainment of the focal organization, (2) perceived ability of the focal organization to deliver its services, (3) perceived receptivity of the client population to the domain claims of the focal organization. These three dimensions coincide with three essential elements of domain claims out- lined by Levine and White. This helped to reduce some of the confusion within the research literature on domain acceptance and domain consensus by clearly specifying three areas that organizations can agree or disagree. The processes by which these perceptions of the focal organization are established and maintained are probably some of the important mechanisms for securing organizational legitimacy. For example, the perception of the focal organization's goal attainment level is probably influenced 135 by such factors as: (1) actual amount of goal attainment, (2) effectiveness of formal publication of goal attainment (i.e., local newspaper articles, WIN newsletter which is circulated throughout the state with personalized accounts of program successes, publication of failures by welfare rights organizations, etc.). The perception of the focal organization to deliver its services may be more influenced by such variables as: (l) "indeterminant technology" a technology of low or unknown effectiveness (Hollister, 1970; Thompson, 1967: 134); (2) convincing "self-validating "1 (Hasenfeld, 1972: 261): (3) the amount of ideologies exchange of Opinions and ideas between the two organizations. Finally, the perception of clients receptivity to the domain claims of the focal organization may be strongly influenced by the amount of actual contact and rapport the set member employees have with the clients being transferred to the focal organization. One of the needs of future research is to explore the processes by which these determinants of domain acceptance 1Hasenfeld defined self-validating ideologies as "internally generated belief systems about an organization's effectiveness that negate assessment criteria used by external groups, or those based on evaluation of outcomes" (Hasenfeld, 1972: 261). Such empirically unsubstantiated positive perceptions of effectiveness minimize the organization's dependence on the external environment for domain validation. These ideologies, if convincing enough to the set member, may be used as the basis for assessing the focal organization's ability to deliver its services. 136 are established and maintained since once these perceptions have been formulated the acceptance or rejection of domain claims results with no small consequences to the focal organization and its clients. Relatively little research has been conducted on the various mechanisms used to obtain acceptance of domain claims by set members (Perrow, 1961; Ellig and Halebsky, 1961). Perrow, for example, suggested that prestige of an organiza- tion may be important in Obtaining acceptance of organizational goals and new resources. Perrow (1961) illustrated some of the functions and dysfunctions of organizational prestige of a 300 bed voluntary general hospital. He noted that the basis of much prestige may be either the quality of the goods or services produced by the organization or some value laden symbol (i.e., museum of medicine, physical appearance of the building where the product is made, organization's labor policies, political leanings, charitable donations). The former set of criteria Perrow labelled "intrinsic" and the latter set "extrinsic" characteristics. To promote intrinsic factors as the basis of prestige the hospital had to use different strategies than to promote extrinsic factors. In order to develop prestige based on intrinsic factors the hopsital used: (1) "validating groups (i.e., research grants for medical research from well-reputed agencies such as the American Cancer Foundation) and (2) indirect indexes (i.e., reputation of personnel, specialized equipment). To promote extrinsic factors as the basis of 137 prestige, a number of advertising techniques were employed: (1) human interest stories in local newspaper, (2) publicizing the availability of "hotel" characteristics such as wine with the meals and beauty salon for maternity patients, (3) circulating coffee carts for visitors, (4) television sets for patients. Perrow further investigated the effect that the promotion FA of each of these two sets of factors had on specific audiences P which in some way added resources to the hospital. For example, local donors contributed $100,000 in new facilities (i.e., extrinsic characteristic) and expected to have these facilities publicized in national magazines. The national research agencies who sponsored a considerable amount of research projects at this hospital viewed such publicity negatively as they stressed the results of research (i.e., intrinsic characteristic) and not the means. Future research on the basis of prestige and consequences of it for acceptance of domain claims, transfers of resources and eventual goal attainment would be a valuable asset to our meager understanding of the processes by which the environ- ment influences the legitimation of an organization. The other set member variable (i.e., salience of the focal organization as a reference group) and the single focal organization variable (i.e., internal consensus) were 22p found to be statistically predictive of domain acceptance. These findings suggested that these variables were not very helpful in explaining the development of domain acceptance. 138 Reference group theory states that if a group is perceived as important for the self-concept of the individual then the goals of that group would probably be accepted. However, in this case the legal requirements to interact with a specific organization's employees on a professional basis may have resulted in the creation of a "forced" reference group whose good opinion of you is considered an important FF asset but whose goals need not be accepted, especially in E light of low evaluations of the focal organization. More- 1 over, in our study, reference group salience of the represent- ’§ atives of the focal organization was also low relative to the :3 other available sources of identification for social workers. More important than the failure of reference group identification with the focal organization to predict acceptance of their domain claims is the finding that high co- Operation and support as well as high exchanges of ideas, Opinions and information also did pep correlate with domain acceptance (i.e., see Table 20). The frequently alleged positive value attached to these activities by many inter- organizational analysts, as noted by Hall (1973: 3), was not strongly substantiated in this study. The correlations between these variables and domain acceptance were positive but low. Such weak support for the culturally prescribed adherence to co-operation and support and frequency of inter- action should caution prOponents of this strategy for inter- organizational relations against a unilateral acceptance of all] ]:).\' 139 this approach. Some of the dysfunctions of too much co- operation and exchanges of ideas should also be recognized for a more balanced view of the effective form§_of inter- organizational relations. Peter Blau, for example, considers both the positive and the negative consequences of co-Operation and the free flow of ideas for intra-organizational relations. Blau recognizes some of the dysfunctions of such behavior for social work agencies: "By providing social support, challenging stimulation, error correction and laissez- faire competition for respect among partic- ipants the free flow of communication contributes to finding solutions to problems, to making decisions and to creative thinking. But the battleground of ideas generated by a free flow makes co-ordination more difficult.” (Blau and Scott, 1962: 139) At the inter-organizational level of analysis two researchers have recently questioned some of the alleged positive value of high co-operation and exchanges between organizations. William Reid has stated some of the reasons for the absence of interagency coordination in delinquency prevention and control on the basis that such coordination may not be positively functional for the agencies involved although it may be the most effective means of servicing the needs of the community. Reid summarized his position in the following quote: "What may be new - at least to some - is the notion that co—ordination may not necessarily be advanteagous to the agencies involved, as they define their ‘goals and functions. We have often tended to view co-ordination in terms of what is desirable or functional for 140 the community. From the community's point of view, co-ordination of services is a cherished objective. Agencies may also desire this Objective in the abstract, or for other agencies, but may not be committed to improving their own levels of co-ordination. What may be functional for the community may not be functional for the agency." (Reid, 1964: 427) The results of our research suggest that high co- operation and high transfers of resources may not only be ‘7 dysfunctional for the agencies involved but also for the communities involved since under such conditions in the WIN program the client and the community both suffer as a result. ‘111' 1"? Richard Hall (1973) has recently reassessed the assumed positive relationships between co-operation, frequency of information exchanges and good communication. The results of his study of eight agencies focusing on "youth problems" (i.e., juvenile center, juvenile police) indicated that conflict, not co-operation, was positively related to good communication and frequency to contact. Hall deduced that, "What appears to be happening here, is that communication patterns can be of good quality, be frequently used, and be based around conflict situations." (Hall, 1973: 10) In short, at both the intra and inter-organizational levels of analysis there are a number of researchers who are considering the positive and negative consequences of co- operation and frequency of interaction. The strong ideological emphasis on sharing and the assumed "magical" operational relationship (Black and Kase, 1963) between co-Operation and goal attainment has seldom 141 been seriously questioned or empirically tested by either practitioners or researchers until very recently (i.e., Reid, 1964; Hall, 1973). Nonetheless, co-operation, co- ordination and integration remain the utopian biases of social planners as well as exchange theorists. Such strategies are consistently projected as having positive functional consequences for the organizations and clients involved. This research raises some serious doubt about the universal applicability of co-operative strategies. The maintenance of a highly co-operative relationship may be so time comsuming that there would be insufficient time to actually perform the task. Another alternative explanation would be that frequent communication with members of the focal organizatiom may lead the set member employees to negatively evaluate the competency level of the focal organization personnel. In either case, co-operation or frequent contact may lead to low goal attainment. The only focal organization characteristic which was tested as a determinant of domain acceptance was internal consensus -—- the degree of acceptance by members of the focal organization about their own domain claims. This has been assumed to be high by most inter-organizational writers. However, we did not find complete acceptance of domain claims by the members of the WIN teams. High or low internal consensus seemed to have little noticeable effect on whether the employees of the set member eventually accepted the domain claims. The source of domain acceptance did not appear to be l.||llllll ll Ill llli 142 within the attitudes of the focal organization's employees and transferred to the set member employees but rather lodged in the perceptions developed by the set member personnel quite independent of the claims made by the WIN team members. The second part of our study treated domain acceptance as a determinant of resource transfers from the set member to the focal organization. These resources, in turn, were expected to result in high goal attainment for the focal organization. In our research, high domain acceptance led to an appropriate or an effective level of resource transferred for high goal attainment. However, this appropriate level was 223 the highest number of referrals per team. There were one-third less referrals made to the four most successful WIN programs than the four least successful programs. Excessively high referral rates were dysfunctional for program success and were interpreted as indications of program sabotage. Such unnecessary referrals only added further bureaucratic handicaps and hindered the efficient training and placement of WIN clients. Transfer and exchange models of inter-organizational relations should not naively view all exchanges between organizations as either indications of co-operation or automatically positive in their functional consequences. There is a need for future researchers to clearly specify the causal linkages between the number and quality of units being exchanged and their functional and dysfunctional 143 consequences for goal attainment for the organizations and clients involved. Simply exchanging or transferring resources does not "magically“ imply either that such an exchange is beneficial to the receiving organization or that it is designed to be beneficial by the set member who initiated the exchange. Sociologists of knowledge (Berger and Luckmann, 1966) have often remarked on the inevitable cultural impact on the choice of explanatory models used to explain social phenomena. Intra-organizational research has frequently been accused of having a strong managerial bias (Carey, 1967) and generally supportive of the status quo. The predominance of the exchange model in inter-organizational research, once again, indicates the power of the cultural norms (i.e., co- operation and co-ordination) to influence scientific models. The conflict models of inter-organizational relations have been relatively undeveloped and limited to infrequently reported case studies. This neglect of conflict and dominance elements within the social context is, thus, as characteristic of inter-organizational relations as is the general situation of sociological theory and research (Coser, 1956: Dahrendorf, 1958' Wrong, 1961). This research found non-acceptance of domain claims to be the basis of frequent and excessive referrals, which was interpreted as program sabotage since such excessive referrals led to low goal attainment. Conflict and co-operation are both inter-organizational realities. The neglect of the former 144 will only promote a distorted view of inter-organizational relations. There are some researchers who conceive of inter- organizational relations as including both elements of co- operation and conflict (Reid, 1965; Aiken and Hage, 1968; Hall, 1973). The inter-organizational area of sociological interest is young enough that early reCOgnition of the need to develop rm comprehensive models which incorporate both elements of r conflict and co-operation can result in avoiding some of the pitfalls and status quo biases that have plagued sociological theory and research in the past. 5; Another contribution of this research is to indicate the consequences of inter-organizational relations for clients. Klonglan and his colleagues (1973) have noted that there has been almost no attention given to the effects of inter-organizational relationships on clients, especially in the case of service organizations. Our research has attempted to begin the process of indicating how domain acceptance encourages apprOpriate levels of resource transfer and eventual higher client service. Such an approach tries to detail some of the causal linkages between inter-organizational variables and client outcome variables in the manpower development organization set. Finally, this research has found that the three dimensions of domain (i.e., goals, services, population) outlined by Levine and White were helpful in indicating some perceptual determinants Of domain acceptance. These same 145 three factors may also be equally suggestive of a much needed typology of inter-organizational relations (Hall, 1973). Since this area of sociological interest is relatively new the following topology is only suggestive of a general direction that future researchers may find fruitful as a means of arranging many of the divergent studies that characterize much of the research literature. The typology also indicates a direction for future research. Figure 8 presents the typology of inter-organization relations according to the three dimensions of domain acceptance and the consequences for transfer of resources and goal attainment. The typology assumes that all other relevant factors are held constant. One such factor of particular importance is the degree of dependency of the set members on each other for necessary resources for goal attainment. With sufficient funds, equipment and clients supplied from outside the set the importance of domain acceptance is seriously reduced as is the need to exchange of transfer resources (Levine and White, 1961; Litwak and Hylton, 1962). Under such conditions the predictive power of this typology is diminished. Under the opposite conditions of high interdependency of set members, however, the typology should be most appropriate. A second limiting variable is the function being performed by the individual agencies in a set. This restricts the kind and amount of resources which will be transferred. Similar kinds of organizations (i.e., treatment hospitals) 146 usoacamuu< Hwoo poo moouoomom mo Hmmmcmue you moonosaomsoo on» com mosmuaooo< onaon mo msowmomawn ou wswpuooo< maOHuoHom HOOOHOONHomquIHOOOH mo awoaoomell.w ouowam 30H oumumooa oumumooa Ouwuopoa oumuopoa oumumooa oumuopoa :35 onH Imu .U woos no Hmow you gone Ho>oa Qmmmmmmzumm Ho>HHon cu NDHHHB< tam moow>uom mo moomummoo< .N 36H 36H 36H 36H ean aan emHs :wHe mHmoo mo mosmuomoo< .H "muzdammoo< zH