Collaboration in social welfare A framework for analysis Bob Hudson Collaboration is a paradoxical concept in the field of social welfare. There can be little doubt that the notion is in vogue. The desirability of some form of collaborative activity has become a sine qua non of effective practice Within the welfare professions, both at .practitioner and policy-making levels. However we know remarkably little about how čpllabórative activity works, why it may initially be developed, how it may be measured ■ or even how it may be defined. Social science research on organisations has tended to be principally concerned with /«^-organisational phenomena. Psychologists have studied the individual in an organisation; social psychologists have focused on the relations amongst members of a ; group in an organisation, and the impact of a group on the attitudes and behaviour of group members; sociologists have studied informal groups, formal sub-units and the structural attributes of an organisation. There has been relatively little attempt to focus on wzřer-organisatíonal behaviour, particularly in the application to welfare policies in Britain. This paper attempts to outline a framework for the analysis of collaborative ačth/iŕy, which may then be applied to a variety of welfare settings. The significance of inter-organisational behaviour : At abroad level, the significance of interlocking networks of organisations has been ; established at least since the publication of C. Wright Mills' The Power Elite in 1956, andit is now more widely accepted that it is interlocking organisations rather than individuals that are at the centre of power systems. Consumers or clients of welfare ^organisations are usually served, processed, ch'anged or harassed not by a single •organisation but by a number of related organisations. From B.Hudson, 'Collaboration in social welfare:;aframework for analysis', Policy and Politics, I5^987ypp;'l75-S2. ' -r^^ĚSm ' '34Ö 4&m OLLABORATION IN SQGI^^^^j!^^p^K^;- ' '341"" One approach to this difficulty wóuíďbeto encourage awareness on the part of the elated organisations of the impact of their activities upon the individual client or onsumer. This is premised upon certain assumptions about the motivation of inncipal decision-makers, most notably that a prevailing spirit of altruism will result in o-operation as soon as individual or community needs become known. However, it aay be more realistic to assume not only that inter-organisational collaboration in ocial welfare has no qualities of spontaneous growth or self-perpetuation, but also hat organisations strive to maintain their autonomy. From an agency's viewpoint, collaborative activity raises two main difficulties. First, Lloses some of its freedom to act independently, when it would prefer to maintain ontrol over its domain and affairs. Second, it must invest scarce resources and energy n developing and maintaining relationships with other organisations, when the poten-lal returns on this investment are often unclear or intangible. Hence it could be losited that an agency prefers «of to become involved in inter-organisational relation-hips unless it is compelled to do so, and that simple appeals to client well-being may onstitute an insufficient motivation. A much more powerful motivation concerns the realisation of organisational goals. Velfare organisations do not normally possess or control the entire complement of esources needed for their goal accomplishment. Funds, facilities, personnel or other esources may be lacking in some measure, and organisations may therefore enter into xchanges with one another to acquire needed resources. Yuchtman and Seashore 1967) go so far as to define 'organisational effectiveness' as 'the ability of an organi-ation to exploit its environment in obtaining resources, while at the same time naintaining an autonomous bargaining position'. In Britain, there is a small but growing empirical literature on collaborative welfare ictiviry Much of this stems from 'community care' policies and focuses in particular tpon the relationship between health and personal social services authorities, but the ollaborative theme has also been scrutinised in other diverse fields such as child ibuse, the treatment of juvenile offenders and the housing/social services interface. Yhat has been noticeably absent is a theoretical framework to bring some order to the hidings Transfers of resources among profit-making organisations are mediated by noney and behaviour may be explained in terms of economic theory and market lehaviour, but these concepts do not appear suitable for explaining interaction unongst welfare organisations. This chapter will attempt to provide a more appro-mate framework. Modes of organisational co-existence •Tot all organisational relationships are conducive to collaboration. Litwak and Hylton 1962) identify three modes of organisational coexistence, each with different mphcations for the likely development of collaboration •*■ independence, interdepen-lence and conflict Two orgamsationsrnay be said to hemdependentof oneanotherif neither needs .the 342 ORGANIZATIONS^ RULE'S AND BUREAUCRATS other's resources to accomplish its goals, and neither is interfering in the other's goal achievement. When this occurs, there is lit :le need for exchange between organisations, and since agencies do not generally co-operate unless they have to, one would predict a low level of collaboration. If the presence of collaboration is to be broadly accounted for, then perhaps the most general explanation for its occurrence may be found in the notion of organisational imerdependence, where each organisation perceives that its own goals can be achieved most effectively with the assistance of the resources of the others. : Co/z/frcrarises when the goal achievement of one or more organisations occurs at the expense of the goal achievement of others. JAlthough conflict per se may constrain or block inter-organisational exchange, its presence also indicates some potential for such exchange; therefore, inter-orgam'satiqnal conflicts may be considered a more favourable condition for collaboration thari independence. Organisations may move from interdependence to conflict or vice versa, through relatively subtle shifts in goals or available resources. Molnar and Rogers (1979) further distinguish between structural and operational conflict. Structural conflict occurs over the basic identities and responsibilities that define a relationship, and reflects an inability to establish or maintain the basic rules or principles that govern the relationship. Operating conflict represents the disagreements over task expectations or role performance within an inter-organisational relationship, and can be viewed as a continual process of mutual adjustment between interacting organisations. Given that some scope for collaborative activity seems feasible or actually exists, how can we explain any specific linkages that develop? A threefold focus is available: upon the environmental context; upon the comparative properties of an organisational network; and upon collaborative linkages themselves. These three approaches are not in conflict, but should be viewed as complementary approaches. A total analysis of inter-organisational relations would require a thorough understanding of the interplay between the variables operating on all levels, but such an analysis is dependent upon the delineation and operationalisation of these variables. The environmental context External factors in the environment may create the necessary preconditions for inter-organisational relations by affecting the ability of an organisation to function independently. In 'turbulent fields', individual organisations, however large, cannot expect to adapt successfully, simply through! their own direct actions. Emery and Trist (1965) identify several indicators of 'turbulence' - a field containing a relatively large number of organisations; inability of agencies to satisfy the demand for services; an unstable social situation; a new programme or piece of legislation; a retrenching economy. Much of this could apply to Britain. : './although a turbulent environment establjsliĚs/a context for collaborative decisionmaking andperhaps creates a necessity for äčtprí,íthe tendency of organisations to act as; closed systems^, and resist such:^es|i|igs^ LABORATION INSOCIÁLW;EyF:A^ ,ťs research, it appeared.tháťenv^ were Sreat enough to induce ncies to talk together and entera reiäťively loose confederate arrangement. It was until much later, when the financial rewards for a closer relationship were tripled, t the agencies give up anything of value.Clearly the environment is important, but of itself sufficient, to explain the existence and nature of inter-organisational itionships. More immediate factors require to be examined. I he comparative properties approach 3 comparative dimension has been viewed as an initial determinant of the occur-ce of collaborative activity and as a factor in its continued existence and success. The comparative properties approach involves examining the similarities or differ-es of interacting groups or organisations on certain attributes or dimensions which istitute a set of conditions that continually shape the pattern of interaction. I effect, the comparative properties are seen as independent variables, and any jsequent collaborative mechanism as dependent variables. The literature identifies eral comparative property prerequisites for the creation of collaborative activity. er-organisational homogeneity is variable is concerned with the degree to which the members of an interaction twork exhibit functional and structural similarity. Reid (1969) takes the view that hough similarity of goals is not necessary for collaboration to develop, it is likely to d to additional cohesion, and more extensive and stable exchanges. Some of the -rature on health/personal social services forums would certainly suggest that e obstacle to more fruitful collaboration is the difference in value-systems and als amongst the participants. Whilst all may find virtue in a 'banner' goal oi immunity care', the operationalisation of the goal may reveal conflicting underlying .umptions. Hasenfeld (1972) makes a useful distinction between 'people-processing' and ■ople-changing' organisations. Within the welfare field, the traditional focus has en upon the latter, whose explicit function has been to change client behaviour. 1 lucation and social work organisations are obvious examples. 'People-processing' organisations process people and confer public status upon 2m, and thereby shape people's lives by controlling their access to a range of settings, amples would be Job Centres, Observation and Assessment Centres and Juvenile luxts Teople-processing' organisations are much more likely to be involved in collabor-Lve activity. Their raison d'etre is the classification and disposition of clients, and the actions to the status they confer occur mainly outside their boundaries. Staff activity Jl consist of transactions wife tion-dKppsifiori starüa'^ the chents^'Peop^^ more likely to. be insulated froiř. ;:344: ORGANIZATIONS, RULES AND BUREAUCRATS "hypothesised that 'people-changing' organisations will be less attracted to collabor-=;.ätíV;e activity than 'people-processing' organisations. This notion would help to explain whyi for example, social work has a poor record of 'boundary transaction' activity with organisations concerned with social security and housing. :T)omain consensus There is some overlap between the notions of organisational homogeneity and domain consensus. Any discussion of inter-organisational analysis either explicitly or implicitly deals with the concepts of domain and domain! consensus. Thompson (1967) defines it as: | a set of expectations, both for members of an organisation and for others with whom they interact, about what the other organisation will arid will not do... it provides an image of the organisation's role in a larger system, which in turn serves as a guide for the ordering of action in certain directions and not in others. j Braito et al. (1972) operationalised it as an organisation's statement as to whether or not a particular organisation should be involved in an agency formed for a specific ■ purpose. The establishment of domain consensus requires the resolution of some potentially difficult issues. First it requires agreement on specific organisational goals. Second it assumes a compatibility of organisational goals, philosophies and reference orientations. Finally, it requires some agreement amongst kindred professionals upon their pqsitionin a hierarchy of professionals. Clearly these will not be attained easily. An organisation's history and association with a particular problem should increase the possibility of the problem being within its domain, but even this may be insufficient to maintain the domain consensus. The public,! political or other organisations may become dissatisfied with the progress of a particular organisation, question its domain and withdraw legitimation. The manner in which in the 1940s local government lost its responsibility for both hospitals and poor law relief may be examined within this ■ framework. i The establishment of a consensus over domain has two important implications. The mpst obvious is the securing of legitimated claims. Possession of a domain permits an organisation to operate in a certain sphere, claim support for its activities and define ^rpper practices within its realm. Moreover, the authority to conduct activities is generally assumed to imply a claim upon money adequate to attain performance in a prelcribed sphere. :^vThe other implication is for collaborative conflicting-views on this. When organisations have similar domains, they are likely to beiaware. of one another and to have the resources needed to help each other achieve their:_resp_e_ctive goals, but similar domains also increase the potential for territorial {disputes and-competition.Van de Ven (1976) proposes a concave-shaped relationship between domain s^ :Äŕ i low Romain:; simil arity ŕ areif unJü^h/lirapäh^ tiinterměojMéíT^^ activity itself, but the literature has íiäšiťfcííšíissKiíä-s COLLABORATION IN SOCIAL WELFARE 345 gjjgjjpfc Hft. Ills1' Bil lj§Bpi'. ■l iÜiv! tt- 'n T'r B. i Blips ĺ Hilar jjgjjsjffi IIIilI Uni. gsgsnj IpjJBiiLi -IBB = lliiin lllfli Uff n h Network awareness It has already been suggested that interdependency is the most fertile ground for collaboration, but organisations must have awareness of their interdependence. Not only must they have knowledge of one another's existence, but they must also perceive a possible matching of goals and resources that would result in more effective goal achievement. Collaboration will also be affected by the extent of positive evaluation -the judgement by workers in the organisation of the value of the work of another organisation. This may be affected by the length of time that organisations have been members of a network. A history of distinct administrative divisions, separate patterns of accountability and isolated patterns of training and professional socialisation all militate against positive network awareness. It is not uncommon in the literature to find not only a poor understanding of the roles of related professions and organisations, but also an unduly critical appraisal of any potential contribution. Most of the reports of inquiries into child abuse cases provide ample evidence of this. Organisational exchange The concept of 'exchange' can be traced back to the work of Marcel Mauss, who proposed an explanation for the seemingly one-way transfer of resources by suggesting that receipt of the gift created an obligation in the receipt. The concept, therefore, implies that no goods or services are ever transferred without reciprocity of some kind being involved. Theorists have utilised exchange notions to provide a loose conceptual framework for their analyses, but few attempts have been made to apply these to inter-organisational relationships in welfare. In their seminal contribution, Levine and White (1961) defined 'exchange' as 'any voluntary activity between two organisations which has consequences, actual or anticipated, for the realisation of their respective goals and objectives'. Cook (1977) points out that the problem with this definition is that it incorporates any form of voluntary activity, thereby rendering the term synonymous with interaction. She prefers to confine the concept to those situations where interactions are based upon reciprocal reinforcement - where exchange provides ioreach actor a reduction in organisational uncertainty. An exchange analysis is fruitful because it focuses attention upon power processes, which are fundamental to an understanding of collaboration. In exchange relationships, power is linked to dependence. It is precisely because the needs of both participating parties need to be fulfilled by an exchange (i.e. it must be beneficial to both) that an integrated and rational system does not always evolve. There are many cases where a transfer of resources from one organisation to another may be desirable from the viewpoint of a co-ordinated system, but may be beneficial to only one party. The reluctance of the NHS to enter into unilateral resource transfers to local authorities in pursuit of a community-based scheme for some clients is an illustration of this. An exchangexelation is balanced when the actors have equal power (or equal -levels,:of-dependency), but equality is not\-a.precondifo necessary is tiiat neiüiei^arty^s pp^ ;"346;;.. organizations, rules and bureaucrats WÜlramount to little more than the formalisation of the clear dominance of one partv over: another. I -.'.-■.- :;:",'."■ I ..-"-"-'.■"" ! Alternative resource sources \ Ä;';iinpörtant element in organisational exchange is the availability of alternative sources. rAn organisation is less dependent upon exchange relations with other organisations in its network to the extent that it has accessibility to elements it needs frpmi-outside' sources. In terms of Hasenfeld's (1972) 'people-processing' organisations^ the availability of such alternatives will i increase the discretion of personnel responsible for processing and reduce the pressure to form links with any one outlet. Without such alternatives, the organisation would be under pressure to control either itsihtake or its output of clients, or both. It may, for example, only accept clients for whom disposition resources are available, or iti may confer a 'holding' status, linking theperson to an outlet but defining him or her as unready for disposition. Hasenfeld's study of employment placement agencies found that clients for whom job opportunities were not available were more likely to be classified as requiring counselling than similar clients for whom job opportunities were available. In Britain, this framework could help to explain the functioning of observation and assessment centres, the availability of assessments and statements under the 1981 Education Act, and the continued high use of custodial options by juvenile courts. : In the comparative properties approach, thejbasic unit of analysis is the network of organisations, consisting of a number of distinguishable organisations having a significant amount of interaction with each other. Such interaction may at one extreme include extensive reciprocal exchanges, and at the other intense hostility or conflict. [The] variables [...] above may help us to explain such variations. Benson (1975) puts forwartl the notion of 'inter-organisational equilibrium1. An inter-organisational network jis said to be equilibrated to the extent that participant organisations are engaged in Highly co-ordinated, co-operative interactions based upon normative consensus and I mutual respect. He hypothesises that_ ; thěreis a tendency towards equilibrium, and argues that increases in one equilibrium component (partly covering [the variables] above) will tend to be associated with increases in the others. Similarly, decreases in one will be associated with decreases in the others. The focus on collaborative linkages The third and final focus is upon the actual dimensions of interaction or exchange between organisations. Four key dimensions tend to be used for examining linkage mechanisms. J^ěgreeqffórntalisation -Several;analyses have noted variations:in:ffie;e^e^;tp^vyhich the requirements and ^chäŕachjÄ aretwo^main ways" OLLABORATION IN SOCIAL WELFARE 347 n which we can attempt to assess the degree of formalisation of a relationship - the ĺ xistence of administrative or legislative sanction, and the existence or otherwise of an itermediary co-ordinating body. The extent to which an interdependent is given official sanction by the parties nvolved refers to the degree to which rules, policies and procedures govern inter-gency agreements and contracts. In social welfare settings, informal tacit arrangements occur quite frequently among organisations - cases may be referred from one gency to another and ideas can be exchanged - but formal agreements are less ommon. This may be in part attributable to the greater commitment required of a formal agreement, and the potential threat to organisational autonomy which it poses Formalisation increases as an agreement is verbalised, written down, contractual md, ultimately, mandatory. Mandated interactions involve laws or regulations pecifying areas of domain, information and financial obligations, the most obvious orm being one externally imposed by the legal or political system. Aldrich (1976) ound that these tended to be more intense, unbalanced in favour of one of the organisations studied and associated with lower perceived co-operation. Benson (1975) hypothesised that in voluntary, non-mandated situations, domain onsensus and positive evaluation were preconditions for collaboration, and the >rganisations attempted to exert power as the exchange occurred. However, when ollaboration is mandated by law, the roles of the interacting organisations have üready been defined, and domain consensus should not be an issue. In such situations, positive evaluation becomes a key issue for the interacting organisations. In the operation of joint finance, for example, a form of mandated elationship exists, but there may continue to be professional and organisational jkirmishing over the respective abilities and perspectives of health and personal social services The second dimension of formalisation is the existence of an intermediary coordinating body, which offers a measure of structural as apposed to agreement formalisation. An agency may be considered 'co-ordinating' if one of its important functions is to bring about exchanges amongst other organisations. Reid (1969) distinguishes between two major strategies that may be used to achieve this - the facilitation and induction of interdependence. The 'facilitation' of interdependence rests upon the assumption that the organisations to be co-ordinated are already close to interdependence or are ready to move m this direction. The degree of facilitation required will depend upon the degree of interdependence already present. At its simplest, this strategy could be executed by the development of inter-organisational awareness of potential interdependencies in relation to existing goals and resources. Amore common and problematic circumstance arises when a co-ordinating agency is confronted with either a high degreéofindependence or conflict among constituent organisations.-:-.; ;\^:>:"::-:--;r:^-?^^r.:'^y;./ This mayáreqmr^ánteŕdepe^^ major changes in their goE^äÄöli^S^i^irtSS^®^^)1^ nrayihe^geněrated/by^the use-of 348 ORGANIZATIONS, RULES AND BUREAUCRA" .resources which the co-ordinating agency possesses or controls and which otht ^organisations desire. In this way, for example, the Housing Corporation may be said 1 .-have brought a degree of order to the voluntary sector of housing. Alternative! inducement may be accomplished through the use of'power' or 'influence', wheret -Organisational goals are modified by external influences. For example, in the 1970s th : BfjSS and Department of the Environment attempted, by the use of Circular (18/7* : and then legislation (1977 Homeless Persons Act), to impose collaborative activii : upon local authority housing and social services departments. Such an approach iue be tenuous and subject to rupture. Degree of intensity The level of intensity indicates the amount of investment an organisation has in i relations with other organisations. Aldrich (1979) identified two measures of intensit First, the amount of resources involved in a relationship. Among welfare agencies, th number of services, referrals and staff support provided to another organisation ai common indicators of the intensity of arelatjon, but measurement is much easier in th profit-orientated sector. Second, the frequency of interaction between organisation although cognisance needs to be taken of the nature of contacts and the authority levt at which they take place. I Most social services contacts involve boundary-spanning personnel who arrange th referral of clients, but these contacts may nót be critical to organisational survival. Rogers (1974) has attempted to develop á scale of intensity in inter-organisation;' relationships, ranging through director acquaintance, director interaction, inform; tion exchange, resource exchange, overlapping membership and written agreement: Ad hoc case co-ordination draws only minimally upon the resources of an organisatior whereas some form of 'programme co-ordination' encompasses a much larger portioi Unless the success of a venture has been clearly established, organisations will b inclined to choose the less intense situation over that which is highly demanding. i Degree of reciprocity \ The reciprocal dimension to a collaborative relationship is based upon the notion c 'exchange', discussed earlier. Interactions need not necessarily be symmetrical - som parties to an exchange may have greater influence in determining the bases an i conditions of the activity than others. | There are several elements of reciprocation, of which the most common is resourc reciprocity- the extent to which the resources in a transaction (or in the longer run, relationship) flow to both parties equally or benefit one of them unilaterally The unii of value transacted between agencies may encompass money, physical facilities an materials, and client referrals. I Information flows are also important— messages or communications about the uml of exchange or the nature of the relationship. Resource and information flow constitute the bases of a social action system, but measurement within welfare setting is problematic because such transactions cannot be assigned an unambiguous valuť BORATION IN SOCIAL WELFARE 34í I : even more true of an intangible such as prestige - the judgement by workers ir i rganisation of the value of the work of another organisation. Indeed, where such 1 ments are negative, self-validating ideologies may be internally generated which i: the assessment criteria used by external groups. Collaboration will be unlikely i b circumstances. » ■ elated notion is definitional reciprocity - the extent to which the terms of e i iction are mutually agreed upon, with equal contributions from all participants i ction is not limited to situations in which both parties set the terms of the nent. Many welfare agencies must follow regulatory or legislative guidelines ir i I lishing collaborative relationships, therefore many of the terms are pre-set ver, an organisation may be predisposed to joint activity, but avoid a specific inter because there is no give and take in problem definition. i of standardisation i represents another dimension taken from traditional models of bureaucratic "are. Just as internal standardisation of procedures smooths bureaucratic i tions and promotes efficiency, so the standardisation of external relations Í; i 11 by cost-conscious administrators. Two aspects of standardisation can be - juished. I it, unit standardisation - the extent of similarity between the individual units o: i sources in a transaction. Litwak and Hylton (1962) hypothesise that the ii ution of formal collaborative mechanisms requires some reliable determinatior II units of exchange and the repetition of the exchange. The argument implies tha I n the exchange elements are ill-defined and shifting, interaction is possible bu 1 interaction unlikely. Second, procedural standardisation ~ the degree o: 111 rity over time in the procedures used for inter-organisational transactions. The i i lity of interaction will have a positive effect on standardisation, because the largei i- investment organisations have in each other, the more they are pushed intc 1 ardised modes of interaction for protecting their investments. Si ndardisation does differ from formalisation. The latter refers to the extent tc II h there is an official agreement, but does not necessitate explicitness on details i -reement that a relationship will be established may be found, but that agreemen 1 save open the nature and operation of the exchange. Vit all of the variables [above] are likely to be of equal significance. The resource in itor of intensity is particularly sensitive and likely to place constraints upon the lu' s. In a situation involving large resources, one could anticipate formal agree--1 ;, standardisation of both units and procedures and reciprocal flow. * 1 rrett (1971) proposes two inter-organisational models. The first is characterisec 3W degree of formalisation, standardisation and intensity and is represented bj *. ind of interaction that occurs among social welfare personnel over referrals. The id is highly formalised, standardised and intense, and is far less likely because ŕ I 'es the kind of investments and commitments which organisations may not; be I i ed to make. Marrett argues that research is needed^n'ö£sp-;m^ m i 1 as on the constraints totl^Tealijation of the^čondij-gi^ 350 ORGA!íš;tZ^tJ|P;NŠ,..IÍU-LES AND BUREAUCRATS Social change in inter-organisational networks Given the preceding analysis, how might inter-organisational networks change in such a way as to promote collaboration? Benson (1975) identifies three sources of pressure. Co-operative strategies Change is sought through agreements and joint planning in which each affected social unit participates and exercises options, and any resultant network alterations are typically compromises agreed upon by thejaffected organisations. Such compromises will usually, involve a process of negotiation and exchange through which each parry voluntarily relinquishes some valued condition in exchange for similar concessions on the part of others. However, despite the conditions for their success are restricted, frequency of co-operative strategies, the since each party must hold something of value for the other party and be capable of resisting the other's demands. Agreements . may cover a wide variety of products or behaviour - exchanges of funds, personnel, facilities and clients are among the most obvious, but agreements to cease disruptive and harassing activities may be of equal importance. Incentives strategies ! This constitutes the purposeful alteration of environmental constraints which may be inhibiting collaborative activity, but falls jshort of a directive. It may be seen as analogous to government regulation of the economy through manipulation of interest rates, tax rates and the money supply. Such a strategy may involve an alteration of the total volume of resources flowing into a network (as in the case of joint finance) or an alteration of resource channels as a means of changing priorities. Tactics such as these typically belong to an executive office or a legislative body. Authoritative strategies --_ Organisations with common vertical ties may be directed to engage in joint activities -- that would not ordinarily occur on a voluntary basis. Relations between agencies may be precisely specified, covering the regulation of contacts, referrals, resource sharing and so forth. This is premised upon the possibility of an executive or legislative body .utilising a dominant position in the flow of resources to specify the nature of programmes and linkages at subordinate levels!, and not merely to encourage or reward such activities. This could cover the example of a local authority taking action to ■prevent 'departmentalism', as in the cases of corporate planning and the fusion of social services and housing departments in several of the London boroughs. It would ■\ also encompass attempts at national government level to rearrange boundaries in the ■pursuit of collaboration, such as the attempt to produce coterminous boundaries :)betweeii health and social services authorities in the 1974 NHS reorganisation. ';■;-^Broadly, such a strategy could cover: •;l;:ithe introduction of new programmes,of iágencies; 2^the fbrmalisation of linkages whiphhaoíliífchěrto been informal or variable; '■■;-":-.. r3^-; (lie rearrangement of an ent^ COLLABORATION IN ^OCY^^^-^^^^^^- ' 351 Conclusion: a warning \r:s-'-(.-it^fi:' This chapter has drawn upon American literature on management and the sociology o: oiganisations to suggest an analytical framework for the analysis of collaborativt activity in social welfare settings. The British approach to this important area has beer unduly empirical. The usefulness of such a framework remains to be put to the test ir a variety of settings. [...] However, a provisional warning should be given against thinking that collaboratior is likely to be a significant factor in resolving welfare dilemmas. Davidson (1976) ha: noted the paradox whereby: commentators on the effects of co-ordination are almost uniformly pessimistic, yet co ordination continues to be promoted as a means of pravidinggreaterrationality in the deliver of services. Warren et al. (1974) have also pointed to the frequent failure of structural arrange ments for collaboration in attaining tangible results. They found that out of a total o 40fi reports of structured co-ordination, only 125 instances of positive tangible result: were obtained, and 29 of these were simply tautological references to improvec communications. The state of the art in assessing the feasibility of successful collaboration is not wel enough developed to include precise operationalised measures for the variable: involved, and further research is needed on operationalisation of the concepts and or testing the relationships between them. But at the end of the day, the fashioning o collaborative relationships of substance remains a job for talented practitioners. References Aldnch, H. (1976), 'Resource dependence and interorganisational relationships', Administra lion and Society, Volume 7. Aldnch, H. (1979), Organisations and Environments, Prentice Hall. Benson, J. IC (1975), "The interorganisational network as a political economy, Administratis Science Quarterly, Volume 20, June. Biaito, R., Paulson, S. and Klongon, G. (1972), 'Domain consensus; a key variable in inter organisational analysis', in Brinkerhoff, M. and Kunz, P., Complex Organisations and thei Environments, Wm. C. Brown. Cook, K. (1977), 'Exchange and power in networks of interorganisational relations'.Socio/ogzco Quarterly, Volume 18. Davidson, S. M. (1976), 'Planning and co-ordination of social services in multi-orgam'sationa contexts', Social Services Review, March. Emery, F. and Trist, E. (1965), 'The causal texture of organisational environments', Huma\ Relations, 18. Uasenfeld, Y. (1972),Teqplé processuig organisations: an exchange approach', America) Sociological Review^ Y^t^^M^k-^]-'' Levine, S. anď/Vv^ite^aí^ f°r the study o ■ intěrorganisa^óna^e^^