WELFARE STATE

Planning

The technology of social work

Jimmy Algie

C. P. Snow's ' two cultures' debate is likely to figure more prominently in the welfare arena in Britain. DHSS is seriously considering applications of systems technology to welfare planning and many professional social workers are trying to master this science-derived approach as a new basis for social work practice. In USA, where systems technologists have worked in social planning for a decade, the debate has begun in earnest.

It was widely hoped that ramifying social spin-offs from technical innovations which the space programme ushered in would stimulate comparable development of an effective social technology. These hopes largely rest on applying systems technology -foundation of the space programme — to welfare operations. As aerospace industries are run down, results of ten years" laboratory ' work in welfare systems technology are tested in first field applications, sponsored by a federal government desperate for new resolutions of increasingly serious social problems.

Based on cybernetics and management science, the systems approach involves systematically redesigning welfare operations as an interconnected whole. Using sophisticated methods of service planning, development, delivery and evaluation, systems technologists seek to provide an integrated response which will achieve a more effective impact on the changing, complex and interlocking social needs of the community.

Simulation models form a cornerstone of systems design. Functions and decisions of Social Services agencies are mapped together with all feasible variations. The simplest form of mapping is a network diagram, familiar from network analysis. In more complex form, the map is translated • into a mathematical picture of reality, mathematical equations symbolising every agency operation and its interrelationships with all other operations. A dynamic representation of the .totality of changing service operations is producea. As the urban planner's model maps traffic flows through towns, so the social planner's model simulates flows of clients through agencies and outcomes of service.

Once a generic pattern of events is adduced, managers feed through the model specific information on such aspects of their particular agency's work as numbers and types of social problem, degrees of associated social pathology, numbers of clients with manifold characteristics, takeup of services and their differential impacts, service benefits and disbeneflts, numbers of workers with varied skills. We are left with a balance sheet of agency performance covering all variables relevant to daily decisions.

Managers use this model to predict what is likely to happen if the agency continues to function by the operational principles on which the simulation was built. If each variable comes out as forecast, the simula tion is judged to represent agency opera tions adequately. When predictions turn out to be wrong, the model is further ad justed till it represents the real life situation, so that better predictions be made in the next round.

Once the simulation is sufficiently accurate to win decision-makers' confidence, experimentation begins. Instead of painfully — often disastrously — experimenting with real life situations as ob liged to at present, decision-makers ex periment on the model instead. They test out in advance probable consequences of decisions: of changing resource allocations between services, of changing service delivery patterns, of changing the mix of staff skills through training programmes, of changing priorities amongst client groups, and so on. What happen g in the model will probably occur in real life if the decision is implemented. We begin to predict precise effects of alternative actions, and make decisions accordingly.

As we simulate changing functions of Social Services_ agencies, so we may simulate changing community_social struc tures, together with .consequential needs and problems. Having simulated The opera tion of agency services and the community social situation, we _join the two models to determine what impacts agency services make (or might make) on community needs, thus simulating interactions between agencies and communities, between workers and clients.

We may conduct simulations at different levels. At the most localised level, simula tion is built around the two key variables: individual client (having certain charac teristics)with individual worker (having certain skills). This kind of micro-simulation model (like HAWSIM first created by Eicker and Bremner at Serendipity Ltd) is a powerful tool for planning and decision-making at field level and will probably be applied first to social Rather more comprehensive are models like DESIM developed by Joshi and Wulff for Applied Human Service Seystems Projects. The first DESIM simulations are being run in Roxbury — Boston's most depressed enclave. The key variables used in DESIM are agency services (having certain degrees of impact on social problems), interacting with social problems (having certain degrees of susceptibility to treatment by these services).

Most comprehensive of all are models like DYNAMO made famous by Forrester working with MIT and Club of Rome futurologists. When used to stimulate the total world situation, these macromodels formed a major stimulus of international concern about depleting natural resources, exploding populations and polluted environments. Less ambitiously and possibly more practically, Forrester's staff use them to map interactions between the total community situation in Lowell, Massachusetts and its social service institutions.

To operate at this more sophisticated level of decision-making, new types of information are required more specifically geared to systematic social planning. Another essential ingredient is therefore an integrated client information system which feeds through information relevant for daily decisions and for updating the simulation models. The models mentioned use Kromholz's CIS which already makes recent British Social Services information systems look distinctly amateur.

A basic systems proposition is that services should be planned and managed holistically, not as unco-ordinated fragments: that social problems should be treated as interlocking clusters, not as isolated symptons. This approach is at variance with the political and administrative set-up of American social service systems. These are characterised by multifarious, dispersed, freelance agencies waging internecine competition for clients, and together with communities increasingly atomised by ethnic discord. The Enlarged Services Bill currently before Congress suggests Seebohm-style organisational integration. Given the difficulties of achieving unified accountability in US public services, most observers are pessimistic of the outcome despite experiments in Michigan and Connecticut. Consequently, the Federal Government is endeavouring to use advances in systems theory as levers to achieve structural integration in face of resistances from social planners and service practitioners.

Social planners and systems technologists seem set on a collision course, both groups having a completely different base in respect of knowledge, values, training and styles of work whilst competing for unusually scarce federal funds. Social planners see the least unfeasible way forward for the 'seventies in terms of small-scale, well-verified social experimentation, pragmatically building on any existing opportunities which offer promise of short-term, marginal improvements, supported by improved political skills and strategies for implementing plans and changes. This attitude was prevalent at a major conference on Centrally Planned Change, held in North Carolina in May. Leading social planners gathered to express their despair

following the lost opportunities for social planning (and social planners) of the Kennedy New Deal and the Johnson Great Society programmes in the 'sixties. The Poverty and Model Cities programmes in disarray, and the programmed budgeting experiment having significantly failed, the advent of second generation operations technology embodied in management science and techniques developed by the Chicago school of welfare economists, were regarded too ambitious to get anywhere.

Systems technologists by contrast are daily arguing the case for fundamentally redesigning social services systems to achieve longer-term, larger-scale effectiveness based on applied systems methodology. Against the day When they break through resistance of field agencies jealous of the power each derives from atomised services and of pessimistic social planners weary at the prospect of a whole new science to be mastered, the management scientists are refining their models with new data arriving daily from field experiments and exploring new ways to communicate their sophisticated methodology to non-scientist practitioners in enthusiastically received training sessions A curious irony arises in comparing British and US situations. As New York's noted planner, Ed Logue, concluded after reviewing recent British initiatives: " the dispersed American non-system badly needs structural co-ordination." Organisationally, British Seebohmised Departments form an integrated system, though we lack supporting advances in operating methods to garner full benefits from this structural synthesis. The British experience is probably more relevant in US public services than ever before, while systems technology is more immediately implementable here than there, and would probably yield greater results. Given the greater potential adaptability of the systems approach in our own organisational setting, it would be foolhardy if our planning and research were not to profit from American development work. If as a result, our capability for systematic decision-making improves — if we become better able to reduce the pool and degree of social problems — then political relationships between the community and its social institutions will become more complex. Increasing decision-making expertise and increasing consumer participa• tion require political management skills of a high order if they are to be run together. A development programme is needed which forwards a more systematised, rational decision-making process in social affairs using the skills of systems technologists and which simultaneously, using social planners' skills, advances political processes and interpersonal competence entailed in harnessing major technological innovations of this kind. Such an integrated programme may offer the only effective alternative to the muddles resulting from muddling through.

Jimmy Algie, Management Studies Tutor at the National Institute for Social Work Training, has just returned from three months as Visiting Research Consultant to Applied Human Service Systems Projects at Brandeis University, Massachusetts.