Skip to main content Accessibility help

Territorial Need Indicators: A New Approach Part I*

  • A. C. Bebbington and Bleddyn Davies


Territorial indicators of need, describing variations in the characteristics of areas ranging from wards to standard regions of the United Kingdom, represent a mainstream application of social indicators in this country. The development of these indicators has, for the most part, been based on an intellectual tradition which has paid little attention to theoretical argument.

In Part I of this article, a typology of existing need indicators is developed. By analysis of some of the best-known and most sophisticated examples, it is illustrated how this lack of theory has severely limited their usefulness in policy practice, particularly with regard to resource allocation, where they are potentially very important. A predominant symptom of the problem encountered with empirically based need indicators is the difficulty of establishing criteria for testing their validity.

For the ‘meaning’ of a need indicator to be clear, the indicator must be theoretically based. More specifically, it should be rooted in theoretical conclusions about the policy of welfare interventions. In Part II of the article, the theory of the need judgement as a cost-benefit decision is used to provide a basis for a need indicator. This method is then explicated with regard to social services provision for the elderly, so as to provide an indicator which is in fact a standard level of expenditure for social services departments in England and Wales.



Hide All

1 Hakim, Catherine, Social Indicators from the Census, office of Population Censuses and Surveys, London, 1975, p. 5 (unpublished).

2 Edwards, John, ‘Social Indicators, Urban Deprivation and Positive Discrimination’, Journal of Social Policy, 4:3 (1975), 281 and 283.

3 Not all territorial indicators are discussed here. It would contribute nothing to our argument to include within the discussion indicators such as those of variations in the incidence of collectively provided outputs produced by Getter and Schumaker, or the territorial incidence of industrial or commercial production per se. See Getter, R. W. and Schumaker, P. D., ‘Contextual Bases of Responsiveness to Citizen Preferences and Group Demands’, Policy and Politics, 6:3 (1978), 249–78.

4 Carlisle, Elaine, “The Conceptual Structure of Social Indicators’, in Shonfield, Andrew and Shaw, Stella (eds), Social Indicators and Social Policy, Heinemann, London, 1972.

5 Edwards, op. cit. p. 277.

6 Hakim, op. cit.

7 Evans, Alan, The Economics of Residential Location, Macmillan, London. 1973.

8 Shevsky, E. and Bell, W., Social Area Analysis, Stanford University Press, Stanford, California, 1955.

9 Wirth, Louis, ‘Urbanism as a Way of Life’, American Journal of Sociology, 44 (1938), 124.

10 Clark, Colin, The Conditions of Economic Progress, Oxford University Press, London, 951.

11 Shevsky and Bell, op. cit. Table II:1. However, it has been argued that the priority accorded to the theoretical base was illusory, that Shevsky and Bell retrospectively developed a theoretical system that would fit both the data analysed in their book in which the system was presented and the results of their earlier purely inductive research. See in particular Robson, B. T., Urban Analysis, Cambridge University Press, London, 1969. More damaging than this are the theoretical criticisms of the argument itself: the theoretical hiatus between a discussion of the social scale and the social differntiation of areas within the city; the exaggerated claims (for instance that their indiators characterized ways of life); and the failure of their scheme to fit the data in subsequent tests. See for instance various articles by Schmid, C. F., with others, in American Sociological Review, 23 (1958); and 25 (1960); Social Forces, 37 (1958); Pacific Sociological Review, 4 (1961); and 5 (1962); see also Hawley, T. H. and Duncan, O. D., ‘Social Area Analysis: A Critical Approach’, Land Economics, 33 (1957), 337–45; Herbert, D. T., ‘Social Area Analysis: A British Study’, Urban Studies, 4 (1967), 4160; Timms, D. W. G., The Urban Mosaic, Cambridge University Press, Cambridge, 1971; and Baldwin, J., ‘Social Area Analysis and Studies of Delinquency’, Social Science Research, 3 (1974), 151–68.

12 The relationship between area analysis and the work of the Chicago school is reviewed in Orleans, P., ‘Robert Park and Social Area Analysis’, Urban Affairs Quarterly, 1 (1966), 519.

13 Thomas, W. I. and Znaniecki, F., The Polish Peasant in Europe and America, Knopf, New York, 1927, p. 1, 129.

14 Unfortunately this argument was underplayed by important textbooks, which associated social disorganization with the loose argument later attacked by Mills, C. Wright in ‘The Professional Ideology of Social Pathologists’, American Journal of Sociology, 49 (1943), 165–80.

15 Thomas and Znaniecki, op. cit.

16 Carey, J. T., Sociology and Public Avoirs, Sage Publications, Beverly Hills, California, 1975. P 107.

17 Ibid. p. 1,023.

18 See for instance Carey, op. cit. p. 66; Small, Albion and Vincent, G. E., Introduction to the Science of Sociology, American Book Corporation, New York, 1894. Albion Small was head of the Chicago school. He had written earlier, ‘There is little likelihood that men who personally observe actual social conditions according to the methods we have proposed instead of speculating about them in study will want to fold their hands and let social evil work out its own salvation.’

19 Burgess, E. W., ‘The Growth of the City’, in R. E. Park, E. W. Burgess and R. D. McKenzie, The City, University of Chicago Press, Chicago, 1925.

20 Robson, op. cit. p. 58.

22 See Jones, D. Caradoc, The Social Survey of Merseyside, Liverpool University Press, Liverpool, 1934; Carr-Saunders, A. M., Jones, D. Caradoc and Moser, C. A., A Survey of Social Conditions in England and Wales, Oxford University Press, Oxford, 1958; and Moser, C. A. and Scott, W., British Towns, Oliver and Boyd, London, 1961.

23 See Webber, R. J., Liverpool Social Area Study, 1971 Data: Final Report, PRAG Technical Paper 14, Centre for Environmental Studies, London, 1975, p. 18. The same argument is used by Webber and Shaw, when they argue that it is important that the process of reducing the large quantity of census information available should ‘contribute the very least possible loss of information, should lose the least pattern…’ – Webber, R. J. and Shaw, M. G., The Social Characteristics of Residential Neighbourhoods: Merseyside, PRAG Technical Paper 24, Centre for Environmental Studies, London, 1978, p. 2. It is for this reason that their analysis forms larger spatial aggregates from smaller areas of similar character: the advantage of enumeration district data is not so much that they provide the researcher with more accurate and detailed information, but that they provide him with much greater flexibility in the choice of the larger units into which they can be aggregated. However, Webber and Shaw did take some account of policy salience in the choice of variables for analysis. Although, within the topic areas, variables were chosen according to the effectiveness with which their distribution differentiated significantly unlike types of area, the authors attempted to choose their forty indicators so as to include all the major topics relevant to strategic planning, including employment; socio-economic status; housing; age and household structure; education; accessibility; migration; and vulnerable minorities. However, the authors emphasized that policy relevance was not a major consideration in the choice of variables for grouping areas into clusters – Webber and Shaw, op. cit. p. 7. The national study of wards and parishes used the same criteria for choosing variables: it sought an approximate balance, first, between the three topics, housing, demographic structure and economic conditions, and, secondly, within these topics, according to the effectiveness with which it had been able to discriminate significantly different types of residential environment during previous similar studies of local areas. Again, the study was at pains to emphasize that significance to policy-making was not an important criterion. The choice was governed by a desire to pick out all the significantly different types of environment thrown up by earlier studies and by the known power of particular census variables to differentiate these from other similar areas – see Webber, R. J., The National Classification of Residential Neighbourhoods: An Introduction to the Classification of Wards and Parishes, PRAG Technical Paper 23, Centre for Environmental Studies, London, 1978, p. 4.

24 Webber, , Liverpool Social Area Study, 1971 Data, pp. 1213.

26 See Hart, D. A., ‘Ordering Change and Changing Orders: A Study of Urban Policy Development’, Policy and Politics, 21:1 (1973), 2742.

27 Webber, , Liverpool Social Area Study, 1971 Data, pp. 1213.

28 Ibid. p. 134.

29 Webber and Shaw, op. cit. p. 12.

30 See Webber and Shaw, op. cit.

31 Compare the ten variables on which Imber based her cluster analysis of social services authorities (Imber, V., A Classification of the English Personal Social Services Authorities, Statistical and Research Report Series 16, Department of Health and Social Security [DHSS], HMSO, London, 1977) with the forty or more used in the PRAG studies. Whereas Webber used the proportion of the population aged sixty-five and over (Webber, , Liverpool Social Area Study, 1971 Data), Imber used the proportion of pensioners living alone, an indicator whose relationship to social services need is clearer; and there are other similar examples.

32 Webber and Shaw, op. cit.

33 Webber, , Liverpool Social Area Study, 1971 Data, p. 14.

34 Webber and Shaw, op. cit. p. 5.

35 Castle, I. M. and Gittus, E., ‘The Distribution of Social Defects in Liverpool’, The Sociological Review, 5 (1957), 4364.

36 Faris, R. E. L., ‘Cultural Isolation and the Schizophrenic Personality’, American Journal of Sociology, 40:2 (1934), 155–64; and Dunham, H. W., ‘The Social Personality of the Catatonic Schizophrenic’, American Journal of Sociology, 49 (1943), 508–18.

37 Caradoc Jones, op. cit.

38 Flynn, M., Flynn, P. and Mellor, M., ‘Social Malaise Research: A Study in Liverpool’, Social Trends, no. 3, Central Statistical Office, HMSO, London, 1972, pp. 4252.

39 Webber and Shaw stated four weaknesses in the derivation of social malaise indicators: first, areas with different types of problem, and thus different needs for intervention, could attain similar scores. The indicators thus obscure the variations. Secondly, the weights attached to individual variables are arbitrary. If they are chosen by principal components analysis they reflect the mix of inputs. Decisions about such a mix could not validly be based on a ‘one-off exercise’, such an exercise making it impossible to evaluate alternative choices of input variables in the selection of priority areas. Anyhow, as Webber argues elsewhere, ‘The technical expert, in selecting the mix of variables for inclusion in the composite index, is usurping a political decision. Even if no decision is politically legitimated, the political choice later appears as technical necessity and obtains a spurious objectivity. The central political decision, the relative importance of different types of deprivation, is therefore unable to be debated’ – Webber, , Liverpool Social Area Study, 1971 Data, p. 13. Thirdly, the implications of some indicators concerning social disadvantage depend upon other factors, which require analysis. Fourthly, this approach must necessarily restrict the analysis to variables which are rendered usable at the enumeration district level by the size of the sample – Webber and Shaw, op. cit. pp. 3–5.

40 See Craig, J. and Driver, A., ‘Identification and Comparison of Small Areas of Adverse Social Conditions’, Applied Statistics 21:1 (1972), 2535. The Home Office circular defined the problem as follows: ‘Areas of acute social need are localised districts within local authority boundaries. They are districts which bear the marks of multiple deprivation which may show itself, for example, by way of notable deficiencies in the physical environment, particularly in housing; overcrowding of houses; family sizes above the average; persistent unemployment; a high proportion of children in trouble or in need of care; or a combination of these’ – Home Office, Circular 225:68, London, 1968.

41 ‘Urban deprivation has never been adequately defined. It is at best an ambiguous term and in all probability its nature and manifestations are constantly shifting’ – Edwards, op. cit. p. 280.

42 The makers of social policy more than those of economic policy justify allocations by arguing a putative connection with highly general goals such as aims for redistribution or social integration, rather than with specific goals. The effects of social services activity on maintaining the ‘integry’ are indirect, obscure and uncertain, and possibly, on occasion, in a direction opposite to what we intend. It is not obvious that a reliance on such rationales will enhance societal commitment to social welfare allocations. Yet, should those who promote such allocations be less enthusiastic in their support of agencies which successfully attain narrow goals than, for instance, the Department of Trade would be in the case of a manufacturer who efficiently made nuts and bolts, who sold a substantial proportion of his output to Saudi Arabia, and who did not fall foul of the government's policies with respect to price and wage increases?

43 Report of the Committee of Enquiry into Local Government Finance, Cmnd 6453, HMSO, London, 1976, p. 220.

44 Ibid. p. 214.

45 Ibid. p. 310.

46 Report of the Royal Commission on Local Government in England 1966–69, Cmnd 4040, HMSO, London, 1969, para. 28.

47 Certainly the evidence against this proposition is very strong – see Davies, B. P., On Local Expenditure and a Standard Level of Service, Appendix 10, in Report of the Committee of Enquiry into Local Government Finance. One cannot but wonder whether, had that committee been differently constituted, it would have reached the same conclusion about the potential of alternative approaches.

48 Report of the Resource Allocation Working Party: Sharing Resources for Health in England, DHSS, HMSO, London, 1976.

49 Compare the further statement in Holtermann, Sally, ‘The Welfare Economics of Priority Area Policies’, Jotirnal of Social Policy, 7:1 (1978), 23–9.

50 The housing investment plan (HIP) is currently an allocation to authorities on a regional basis, each region adopting its own strategy for distribution. Paris, C., ‘HIPs and Housing Need: The Oxford Experience’, Centre for Environmental Studies Review, 5 (1979), 1927, describes the need indicator approach used by the south-east region. The measurement of need is interesting. In contrast to the needs element formula of the rate support grant, which bases it on past expenditure, and the RAWP, which uses mortality as an indicator of morbidity, in the public sector block a subjectively determined points system provides the basis. Paris regards such value-judgements as inevitable and more appropriate than ‘objective’ indicators. Such an approach certainly helps to avoid the problem of the RSG or RAWP methods, namely, that debate about issues of priorities becomes submerged. In reality, however, we feel that it is likely that precedent will establish rules which remove most of the subjectivity from the needs assessment for the purpose of allocation; at the other extreme, there is a risk that the allocation may become excessively volatile from year to year, due to short-run political bargaining. It might be thought that the allocation of resources to universities should be based on a principle of meeting higher education need throughout the United Kingdom as a whole. Yet in practice discussion centres on attempting to combine incentives to efficiency within each institution with a system of meeting the expenditure needs. Hence issues closely resembling those of territorial justice arise regarding the University Grants Committee (UGC) resources allocation, even to the point at which Cook has implied the UGC secretly adopts a needs indicator formula, possibly borrowed from the province of Ontario. This suggestion has elicited a fairly vigorous rebuff. See Cook, W. R., ‘How the UGC Determines Allocations of Recurrent Grants: A Curious Correlation’, Journal of the Royal Statistical Society A, 139 (1976), 374–84; comments on this article in Journal of the Royal Statistical Society, A, 140 (1977), 199209; and a reply by Cook, in Journal of the Royal Statistical Society A, 140 (1977), 511–13.

51 See Office of Health Economics, ‘Scarce Resources in Health Care’, in Milbank Memorial Fund Quarterly, 57 (1979), 265–87; and Drummond, M. F., ‘Sharing Resources for Health in England: The Case of Teaching Hospitals’, in Culyer, A. J. and Wright, K. J. (eds), Economic Aspects of Social Services, Martin Robertson, London, 1978, pp. 141–54.

52 See Buxton, M. J. and Klein, R. E., Allocating Health Resources: A Commentary on the Report of the RAWP Research Paper 3, Royal Commission on the National Health Service, HMSO, London, 1978; Crease, A. L., Darby, S. C., Palmer, S. R. and Patrick, D. L., ‘NHS Priorities and RAWP’, British Medical Journal, 6,149 (1978), 1,4467; Sennand, S. J.Shaw, H., ‘Resource Allocation: Some Problems in Applying the National Formula to Area and District Revenue Allocations’, Journal of Epidemiology and Community Medicine, 32:1 (1978), 22–7.

53 Palmer, S. R., ‘The Use of Mortality Data and Resource Allocation’, in SirBrotherston, John (ed.), Morbidity and its Relationship to Resource Allocation, Welsh Office, Cardiff, 1978, p. 26.

54 Dajda, Richard, ‘Self-Reported Morbidity as an Index of Regional Resource Requirement’, Journal of Epidemiology and Community Health, 33 (1979), 138–41. D. P. Forster has quoted evidence about the comparatively low degree of association between what certain doctors have recalled as the principal condition treated and the underlying cause of death, as well as the sensitivity of resource allocations to small variations in deaths by selected conditions – D. P. Forster, ‘Mortality as an Indicator of Morbidity in Resource Allocation’, in Brotherston, op. cit. pp. 13–24. (In another article the same author showed that there existed only in statistically non-significant rank correlations between standardized mortality ratios and all but one of the indicators that he had computed on morbidity from the General Household Survey – Forster, D. P., ‘Mortality, Morbidity and Resource Allocation, The Lancet, 7 05 1977, 997–9).

55 Alternative measures of mortality are examined by Palmer, S., in ‘Mortality Indices in Resources Allocation’, mimeograph, Department of Community Medicine, St Thomas' Hospital Medical School, London, 1978, and the sensitivity of resource allocations to the use of international classification of diseases (ICD) chapters in RAWP calculations is examined in Palmer, S., ‘The Use of ICD Chapters in RAWP Calculations’, mimeograph, Department of Community Medicine, St Thomas' Hospital Medical School, London, 1978.

56 See Barr, A. and Logan, R. F. L., ‘Policy Alternatives for Resource Allocation’, The Lancet, 7 05 1977, 994–7.

57 National Centre for Health Statistics, Synthetic State Estimates for Disability, Public Health Service (PHS) Publication no. 1,759, U. S. Government Printing Office, Washington, D.C., 1968. For a recent example, see National Centre for Health Statistics, Synthetic Estimations of State Health Characteristics based on the Health Interview Survey, Department of Health, Education, and Welfare Publication no. PHS 78:1,349, U.S. Government Printing Office, Washington, D.C., 1977. The methodology has been strengthened as the theoretical basis of varying estimation procedures has become subject to mathematical investigation – see Purcell, N. and Kish, L., ‘Estimation for Small Domains’, Biometrics, 35 (1979), 365–84; and Holt, D., Smith, T. M. F. and Tamberlin, T. J., ‘A Model-Based Approach to Estimation for Small Sub-Groups of a Population’, Journal of the American Statistical Association, 74 (1979), 405–10. The approach was applied in Britain to the personal social services in Davies, B., ‘An Index of Variation in the “Need” of County Boroughs for Services for the Elderly’, Sociological Review (03 1964), 538.

58 Geary, K., ‘Technical Difficulties in RAWP’, British Medical Journal, 1 (1977), 1,367.

59 See for example Jackson, R. R. P. and Himatsingani, C., ‘Measurement and Evaluation of Health and Personal Social Services for the Elderly’, in Canvin, R. W. and Pearson, N. G. (eds), Needs of tlte Elderly, Institute of Biometry, Publication no. 2, Exeter, 1973. This topic is further reviewed in Part II of the present article.

60 As an illustration of the way in which the RAWP formula adopts a pattern of priorities which tends to support the status quo, Bebbington has shown how, if it were applied to a country consisting of half of England and Wales and half of Thailand, the resulting resource allocation would be nearly three times greater per capita in the English half than in the other – Bebbington, A. C., ‘RAWP: A Suitable Case for Treatment?’, PSSRU, University of Kent, 1979.

61 Davies, B., Social Needs and Resources in Local Services, Michael Joseph, London, 1968. The basic article developing the idea of the needs indicator was published in 1964 –Davies, ‘An Index of the Variations in “Need’ of County Boroughs for Old People's Homes’.

62 Imber, op. cit.

63 Imber includes in the subset of variables on which she bases her classification of local authorities the proportion of the elderly in households consisting of pensioners only. However, this does not catch the incidence of various populations of old people differing in their age and household composition – characteristics strongly correlated with need and demand.

64 Ibid. p. 31.

65 Webber, , Liverpool Social Area Study, 1971 Data, pp. 108–10; and Webber and Shaw, op. cit. p. 19.

66 Imber, op. cit. p. 31.

67 Webber and Shaw make this point to support the derivation of clusters of similar enumeration districts rather than taking as units of analysis larger administrative areas such as wards – Webber and Shaw, op. cit. The argument is clearly far more powerful for major authorities.

68 Devon, he argued, includes Plymouth, a city of some 250,000, with problems similar to those of many of the larger metropolitan districts – its situation is cancelled out by some coastal towns and large rural expanses suffering from problems of depopulation and neglect of the infrastructure (for instance transportation, shopping and basic amenities).

69 DHSS, Local Authority Personal Social Services: Summary of Planning Returns, 1976–7 to 1979–80, London, 1978.

70 Gwynne, D., ‘Norms for the Provision of Social Services’, Social Services Research Group Journal, 6 03 1979, 11. John Townsend, in the same issue of the journal, suggested a second way in which the analyses interpreted as reliable needs analyses will affect policy–making, namely, by offering standards of comparison – Townsend, John, ‘The Uses of the LAPS [Local Authority Planning Statements] Exercise’, Social Services Research Group Journal, 6 03 1979, 4.

71 Bebbington found that at most one-half of the variance in any of the need indicators for the elderly developed in Part II of this article is accountable for by intra-class correlation in the Imber classification – Bebbington, A. C., ‘Comments on A Classification of the English Personal Social Service Authorities’, PSSRU, University of Kent, 1977.

72 See for instance the literature of the measurement of cognitive ability and the associated development of factor analysis in the inter-war years.

73 The studies which we have referred to in this article are the leading ones in this field. We would have been unable to refrain from technical criticisms of many others of their genre.

* This article is in two parts, the second of which will appear in a subsequent issue of this journal.

Research Fellow, Personal Social Services Research Unit (PSSRU), University of Kent at Canterbury.

Director of the PSSRU, and Professor of Social Policy, University of Kebt at Canterbury.


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed