Social Science and Social Policy

Martin I A Bulmer

• Éditeur: Routledge
• Année: 2021
• ISBN:
  9781000508864
• Édition: 1
• Sujet: Social Sciences
• Sous-sujet: Sociology

First published in 1986, Social Science and Social Policy addresses major questions concerned with the social utility of social science. The book is divided into four parts. The first part considers the place of social science in the policy-making process and criticizes the rational model which gives a central place to analysis. In part two, five different methodologies for policy research are considered: the use of continuous surveys, public opinion polls, social indicators, evaluation research and social experimentations and the use of qualitative methods. The advantages and drawbacks of each are considered with extensive use of examples. In the third part, the role of theory is examined. Particular attention is paid to the issue of health inequality. In part four, general questions are raised about the use and abuse of social science, including questions about how it can be most effectively disseminated to make maximum impact.

The book is aimed at a general readership and requires no special methodological expertise. It will appeal particularly to undergraduates and graduate students taking courses in social policy, public policy applied sociology and a range of applied social sciences such as criminology, health studies, education and social work.

II

Types of Research Utilization

4

Types of Research Utilization: an Overview

The application of social science research to social policy inescapably involves considerations of social research methodology. Indeed, the principal claim to expertise of social policy researchers lies in their methodological competence, which is a necessary condition for carrying out effective and utilizable research. Without a sound methodological basis, such research can enjoy little credibility. Assessment of the methodological soundness of a piece of applied research is an essential first step on the road to utilization.

Consider three examples. Starting in 1983, the London social research organization Social and Community Planning Research has mounted an annual survey of British Social Attitudes (Jowell and Airey, 1984; Jowell and Witherspoon, 1985). A random sample of the population are questioned in depth, using reliable and tested survey instruments, about their views upon current social issues. The focus is upon attitudes and values rather than behaviour (which is extensively covered in the government’s General Household Survey). Consider one finding from the survey’s 1984 report, shown in Table 4.1. The outstanding priorities for additional public spending by government were health and education, with help for industry third and housing fourth. These findings are also consistent with other work, in the tradition of subjective social indicators (see ch. 7) which sought to establish whether there existed a clear order of priority preferences for different aspects of social well-being. These showed that in all cases health was ranked first (Hall, 1976; Knox, 1976a; 1978). The British Social Attitudes findings were at odds with the government’s spending priorities at the time, which gave highest priority to defence and the police, which appear sixth and seventh in the list. In interpreting these survey results and judging their relevance and importance for policy debates, one must form some assessment of their reliability and validity. Such attitude statements are particularly likely to fluctuate according to the sponsorship of the survey, the way in which the question is asked and the circumstances under which the survey is carried out. The British Social Attitudes survey is carried out in a scrupulously scientific manner, but not all survey and polling data on social attitudes can be taken at face value. This is discussed further in Chapter 6.

Table 4.1 First and Second Priorities for Extra Government Spending as Seen by the British Public in 1983

Q. Here are some items of government spending. Which of them, if any, would be your highest priority for extra spending, and which next? Please read through the whole list before deciding.
	First priority (%)	First and second priority (%)
Health	37	63
Education	24	50
Help for industry	16	29
Housing	7	20
Social security benefits	6	12
Defence	4	8
Police and prisons	3	8
Roads	2	5
Public transport	1	3
Overseas aid	< 1	1
(Sample size)	(1,761)	(1,761)

Source:Jowell and Airey (1984), pp. 77, 98–9.

In the example in Table 4.1 health came out as the highest priority. There is also considerable social science evidence that the incidence of disease and the provision of health care is unequally distributed (Townsend and Davidson, 1982). In the mid-1970s, the British government was concerned that resources within the National Health Service should be distributed more equitably between health regions to take account of differing patterns of health need in different parts of the country. It set up the Resource Allocation Working Party (RAWP), which reported in 1976. This recommended a formula for reallocating central government expenditure between regions, a major element in which was a measure of the need for health care. The working party was composed chiefly of technical specialists, who examined the available social indicators of health need. They distinguished between inputs to and outputs from the health services. The usual indicators of health provision used in previous work were in fact outputs, the results of the existing allocation of resources, such as doctors per patient, bed utilization rates in hospitals, availability of certain high-technology equipment, and so on. But these were not a measure of the state of health of the population prior to service provision, the ‘input’ to the service. To measure this they used as an indicator of need standardized mortality ratios for each health region in the country. These they considered the most satisfactory general health indicator, and one which reflected the existing regional differentials which their recommended formula was designed to modify. Few commentators upon RAWP have been able to fault this use of the indicator, but it is the king-pin upon which their analysis rests (taking account also of population size and the age and sex structure of the population). Problems of constructing and using social indicators are considered further in Chapter 7.

A quite different example of the importance of methodological assessment is provided by qualitative research upon socially deprived working-class families. Both in North America and Britain there is considerable policy interest in the social situation of such families and the need to frame effective policies to alleviate their problems is widely recognized. First, however, the nature of those problems must be understood. Two theoretical approaches have been widely held. In the United States, theories building upon the ‘culture of poverty’ have been popular, holding that poverty is a way of life, with distinct behavioural and attitudinal characteristics such as a fatalistic, present-oriented immediate gratification pattern, marital instability, a high proportion of one-parent families, and insecure low-paid employment. This theory has also been exposed to severe criticism (cf. Lewis, 1966; Leacock, 1971). In Britain the most important series of comparable studies have been concerned with the ‘cycle of deprivation’, the idea that poverty and disadvantage are passed on from one generation to another from inadequate parents, using inadequate child-rearing practices to bring up emotionally, socially and intellectually deprived children. These children fail at school, become unemployed or at best get unskilled jobs, establish unstable marriages and family lives, live in poor housing, and pass on these disadvantages to their children. This theory was put forward in 1972 by Sir Keith Joseph when Secretary of State for Social Services, and formed the basis of a large research programme by the (UK) Social Science Research Council (Brown and Madge, 1982).

Two qualitative studies which cast considerable doubt on these theories are Carol Stack’s study of social support among black families (1974) and Frank Coffield et al.’s study of four English deprived families (Coffield, Robinson and Sarsby, 1980). Each is based upon a study in depth of a small number of cases, building up a detailed and theoretically informed picture of how they live, and showing that the difficulties they experience in their social circumstances are not explicable in the way that the theories postulate. What confidence can one have in such small-scale research? To what extent can these results be generalized? Would another social scientist who spent a similar amount of time with such families record similar information and produce similar findings (that is, are the results reliable)? These questions are considered further in Chapter 9, but they are another reminder that the methodology of research with potential policy applications has to be scrutinized carefully.

These examples point to the form taken by the second part of this book, an examination of the different types of methodology used in policy research. For it is only if these can be assessed effectively and critically that one can hope to take an informed approach to the social utility of social science. Six main approaches will be considered: (1) continuous social surveys providing behavioural data about the national population; (2) studies of social and political attitudes and their pitfalls, particularly opinion polls; (3) social indicators; (4) social impact assessment; (5) evaluation research and social experimentation; (6) qualitative policy research. This is not an exhaustive list, but it embraces the main types of method used in applied research. This way of distinguishing between styles of research places the main emphasis upon the methodology used.

An earlier classification (Bulmer, 1978, pp. 8–9), which remains useful, distinguished between intelligence and monitoring, basic, strategic and tactical research, and action research. The six approaches discussed in this section can be classified in this way, but it may be more useful to consider two main dimensions of difference. One is whether research is diagnostic or future-oriented. Does it seek to analyse what has gone before, or provide pointers to future action? Continuous surveys aim to do both, as do social indicators. Evaluative research is implicitly highly future-oriented, qualitative research primarily diagnostic. Different objectives are embodied in different types of inquiry. They are not all of a piece.

An even more useful distinction is between descriptive, explanatory and evaluative research. Descriptive research seeks to present a factual account relevant to the determination of policy. It can provide estimates of distributional characteristics of some policy-relevant phenomenon, and monitor changes in those characteristics over time. Both continuous surveys and social indicators may perform this task effectively. Explanatory research seeks to understand or account for the attitudes or behaviour observed, for example to explain why inequalities in health or poverty exist and persist. Evaluative research seeks to test the consequences of the adoption of particular courses of action, for example, to test which different type of penal treatment, or medical treatment, or educational programme, is most effective. Evaluative research, it must be made clear in relation to earlier discussion, does not mean research which is value-laden or prescriptive in intent. It means research designed and carried out in a scientific manner to assess the efficacy of particular policy measures in practice.

The distinction between the three is not an entirely hard and fast one. No piece of empirical research is ever an unvarnished account of ‘the facts’, and the idea of a ‘theory-free’ description is untenable. All social description involves conceptualization and measurement in which classification takes place in terms of more general categories, for example income, or disability or social class (Bulmer, 1982, pp. 51–8, 68–79). Explanatory research may sometimes be evaluative, if it focuses purely upon the variables susceptible to manipulation by policy-makers. In general, however, the threefold distinction is a tenable one.

Applied here to the six types of research methodology, continuous social surveys and studies of social attitudes are usually descriptive, though their results may be subjected to explanatory secondary analysis by others. Social indicators and social impact assessment are descriptive with evaluative implications, which may or may not be spelled out. Evaluative research and social experimentation, as their name implies, are evaluative in intent and effect. Qualitative research is usually desc...

Table des matières

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Copyright
• Table of Contents
• List of Figures and Tables
• Acknowledgements
• Abbreviations
• Preface
• I Social Science and the Policy-Making Process
• II Types of Research Utilization
• III Problems of Analysis and Inference
• IV Policy Implementation
• Bibliography
• Index