Research Policy and Planning: The journal of the Social Services Research Group – Vol 19 (1) 2001
Risk Research in a Risk Society
Alison Shaw, Research Fellow in Health and Social Care, University of the West of England, Bristol; Ian Shaw, Reader in Social Work, Cardiff University School of Social Sciences.
The concept of risk has become increasingly prominent in a range of academic, social and political discourses in recent years. Risk has similarly become the preoccupation of several professional groups, and is part of the commonplace of social work education, conversation and practice. This paper suggests two limitations of most social work discussions of risk: first, that wider debates about risk in society impinge only selectively on social work ‘in house’ preoccupations; and second, that there is limited connection between policy and practice accounts of risk, and risk and research. The authors outline, provisionally and illustratively, the value of drawing connections at both these levels. The first section summarises conceptual and theoretical discussions of risk, suggesting links and gaps between this work and social work writing on risk. The second section illustrates how a wider research awareness of risk provides a strengthened purchase on, a critique of, and commitment to, the social purpose of social work.
Risk is a concept which has received increasing attention, with recurring debates about the riskiness of certain actions, situations, and substances. It has become a major issue for ‘lay’ people, ‘experts’, professional groups, journalists, and government (Gabe, 1995). Douglas and Wildavsky (1982) sum up people’s preoccupation with risk in this way: ‘What are Americans afraid of? Nothing much, really, except the food they eat, the water they drink, the air they breathe, the land they live on, and the energy they use’ (p.10). Policy-making has increasingly centred on developing strategies for effective risk communication in a wide range of areas. Risk assessment and risk management have similarly become the preoccupation of several professional groups, including health, welfare and social workers. Within academic inquiry, interest in risk is reflected in research programmes such as the Economic and Social Research Council’s ‘Risk and Human Behaviour’ programme, and developments such as the Centre for the Analysis of Risk and Regulation at the London School of Economics.
Risk is part of the commonplace of social work education, conversation and practice. Requirements to evidence competence in assessing, responding to and managing risk run through the requirements of social work qualifying programmes and requirements for the postqualifying award in social work (CCETSW, 1995, 1997). Descriptions and prescriptions for good practice in child protection, mental health, disability, older people and offenders insist that an informed understanding and response to issues of risk should characterise care management, specific and comprehensive assessments, inter-professional multi-disciplinarity, and user and carer participation in services.
For the moment we want to draw attention to two limitations of most social work discussions of risk. First, we may fairly imply from the opening paragraph of this paper that the extensive wider debates about risk in society impinge at best only selectively on social work ‘in-house’ preoccupations. Second, there has been little or no connection made between policy and practice accounts of risk, and risk and research (Fraser et al, 1999; White, 1999). We want to outline, provisionally and illustratively, the value of drawing connections at both these levels. In the first part of the paper we summarise discussions regarding the conceptualisation of risk, and developments in theories of risk, especially from the social sciences. We halt this part of the paper with a cautious suggestion of links and gaps between this work and recent social work writing on risk. In the second part of the paper we aim to illustrate how a wider research awareness of risk – as central to the social, political, substantive and methodological contexts of disciplined inquiry – provides a strengthened purchase on, a critique of, and commitment to, the social purpose of social work. We suggest that ideas of risk give shape and texture to research ethics, awareness of the political context for research, the uses of research, and the street-level fashioning of risk strategies in social work teams.
Risk and Society: Problems of conceptualisation
Until the mid 1990s the idea of risk tended to be treated as unproblematic within social work discourse. With the exception of the neglected earlier work of Brearley (Brearley, 1982), the meaning of risk was treated in a fairly taken-for granted manner. Yet some basic distinctions have been accepted. First, the history of its use in the English language can be traced to the French term risque, derived from the Italian risco and riscare, meaning to run into danger (Shorter Oxford Dictionary). In its entrance into English, ‘risk’ suggested both chance and consequences, but particularly emphasised the negative consequences: risk as ‘hazard, danger, chance of harm’ (Alaszewski, 1998: 4). Hence, from its seventeenth century integration into the English language, the verb and noun ‘risk’ have carried the dual ideas of danger (‘to run the risk of something’) and probability (as in its actuarial usage) (Brearley, 1982). Second, the general discrediting of positivist ideas of inquiry, associated with the assumption of the ability of science to fully represent reality, has led to widespread debate about the distinction between the reality and the perception of reality. This has been central to discussions of risk, with ongoing conceptual debates on the objective reality or subjective perception of risk (cf MacDonald and MacDonald, 1999). Third, the literature on risk in social work has for some time made a distinction between risk as carried by the social worker and risk as ascribed to others by the social worker (Brearley, 1982). This is captured in the title of a recent discussion of social work staff – ‘the risks they face and their dangerousness to others’ (Burke, 1999). Until more recently social work discussions of risk have also tended to regard the consolidation of practice based on risk assessment and management as generally benign or at least neutral. More recent discussions have been marked by an increasing problematising of ideas of risk, prompted by the persistent note of disquiet that has surfaced (Kemshall, 1998; Kemshall and Pritchard, 1996; Parsloe, 1999; Parton et al, 1997; Parton 1998).
However, there are remaining ambiguities in meaning and conceptualisation (Hayes, 1992; Wharton, 1992). Earlier meanings of risk incorporated both positive and negative outcomes. However, Douglas (1990) notes how risk has changed its meaning from its origins as a neutral term, and is now used only to refer to undesirable outcomes: ‘the word risk now means danger’ (p.3).
Some writers have differentiated between ‘everyday’ and ‘specialist’ or ‘technical’ uses of the term risk (Alaszewski, 1998; Hansson 1989). In everyday use, ‘risk indicates a threat or a danger to an individual or a group of individuals’ (Alaszewski, 1998: 3). In terms of ‘specialist’ uses, risk discussions take place in multi-disciplinary arenas, and are studied in different ways by diverse groups of academics and professionals. Some writers have attempted to classify the major perspectives, which are seen to be actuarial/statistical; epidemiological; engineering; economic; psychological/psychometric; social theory; and cultural theory (Alaszewski, 1998; Renn, 1992). In the scientific literature, risk is associated with mathematical theories of probability: ‘a combination of the probability, or frequency, of occurrence of a defined hazard and the magnitude of the consequences of the occurrence’ (Warner, 1992 : 4). However, in practice it is often an ambiguous term with varying meanings, e.g. risk as high incidence or statistical frequency, uncertainty of benefit and cost, and fortune or misfortune (Bellaby, 1990). Similarly, within the social sciences, there are a range of conceptualisations: ‘Social scientists have used the word risk to mean a variety of things’ (Heimer, 1988: 493), with resulting variations in preferred methodologies and methods for conducting risk research.
A useful typology of risk terms has been developed by Alaszewski (1998), who outlines a ‘risk iceberg’ where ‘risk forms the tip of an iceberg of related words and terms’ (p.10). These inter-related terms include hazard, danger, harm, safety, vulnerability, dangerousness, blame and accountability. He argues that hazard or danger are key terms underlying the concept of risk, where hazard or danger are defined as ‘a potential threat which can result in harm, loss or negative consequences for an individual or group’ (p.10). Harm is defined as ‘the loss experienced by individuals or groups as the result of specific events or actions’ (p.11). Vulnerability and dangerousness are reciprocal terms, where vulnerability is an individual’s susceptibility to experience harm, and dangerousness is an individual’s likelihood to cause harm. He argues that concerns with threats posed by vulnerable/dangerous people are central driving forces behind social welfare policies designed to manage risks. Safety is another related term, seen to be ‘the absence of harm’ (or) the processes by which harm is prevented and avoided’ (p.12), and he argues that safety has become a major preoccupation within governments and regulation. Finally, blame and accountability are key terms, where allocating blame is central to the identification of risk: ‘Blame is the process of allocating responsibility and punishments when accidents occur. Allocating blame is part of the process of accountability, identifying who is likely to be blamed if things go wrong and allocating responsibility and blame when things do go wrong’ (p.13). These terms are particularly important within cultural theories of risk.
Alaszewski’s ‘risk iceberg’ also contains more technical terms related to risk, including risk assessment, risk communication, risk perception, risk-taking, and risk management. Risk assessment is concerned with identifying the hazards which can cause an accident or disaster. This is particularly evident in health and social welfare where it is part of policy development based on ‘scandals’ that have occurred. Accidents and disasters tend to lead to debates about how information about risk is communicated, and how the public view risk, where the ‘failure to effectively communicate information about risk can have major economic and social consequences’ (Alaszewski, 1998: 15). Finally, risk taking and risk management are inter-linked, where risk-taking is about individual risk decisions, whereas risk management is concerned with collective or institutional responses to risk situations.
Before noting the characteristic uses of risk within the welfare and social work literature, and considering risk issues in social work research, we will briefly outline the ways in which social scientists have approached the study of risk.
Risk in the social sciences: Psychology and risk perception
Gabe (1995) argues that ‘psychologists, like engineers, actuarial scientists and epidemiologists, have generally treated risk as an objective phenomena to be measured and explained’ (p.5-6). Cognitive psychology has dominated research on risk perceptions since the 1970s and 1980s. Key work in this tradition includes Slovic et al (1980) and Tversky and Kahneman (1982). They used quantitative methodology, namely questionnaires and psychometric techniques, to examine differences between lay and expert assessments of risks from certain dangers. The assumption behind this approach is that ‘perceived risk is quantifiable and predictable’ (Slovic et al, 1980: 211), and the focus is on cognitive ‘biases’ that lay people exhibit in their risk assessments. The key argument is that lay people use inferences or ‘heuristics’ to simplify risk judgements, but these judgements are frequently biased and ‘lead to severe and systematic errors’ in risk perception (Tversky and Kahneman, 1982: 3). They concluded that lay people have difficulty in making accurate or rational assessments of risk because they are poor at probabilistic thinking. In contrast, expert judgements of risk are seen as more closely related to statistical frequencies and formal probability calculations. So, psychometric research has focused attention on arguments about gaps between the rational (‘expert’) and intuitive (‘lay’) assessment of risk (Broadbent, 1985).
Gabe (1995) argues that the cognitive approach has been useful because it has helped to clarify the properties people include in their risk judgements. However, it has been criticised for treating risk as a purely objective reality, masking uncertainties in assessment of risk through the use of quantitative methods, and failing to recognise that risks are experienced within specific social and cultural contexts (Nelkin, 1985; Turner and Wynne, 1992).
Cultural theories of risk
Cultural theories attempt to address the wider cultural and political contexts of risk. They originate with anthropological work by Douglas (1966), which was developed further in her joint work with Wildavsky (1982). A political dimension is also included through a focus on the way that cultures use risk as a ‘forensic resource’ to allocate blame (Douglas, 1990, 1992). The central thesis is that societies selectively choose certain risks for attention. Risk is conceptualised not as an objective reality but rather ‘The perception of risk is a social process’ (Douglas and Wildavsky, 1982 : 6). Social and cultural factors are seen to influence what risks are chosen for individual and societal attention, and ‘Each form of social life has its own typical risk portfolio’ (p.8). The key element of this approach is grid/group analysis, a classificatory scheme illustrating how the range of cultural variation in the perception of risk can be reduced to a useful number of cultural types. Grid and group are organisational variables, and Douglas and Wildavsky (1982) propose that four cultural types will result from scoring a society as high or low on each variable: hierarchical; competitive individualist; sectarian; and stratified individualist. This typology of cultures has relevance for risk situations by indicating that the social organisation of each type of cultural system will make its members sensitive to different risk situations, and will lead to the favouring of characteristically different decision-making strategies.
However, there are major criticisms of cultural theories (Renn; 1992; Rayner, 1992). Some have argued that the types are too simplistic, masking complexities and variations: ‘Many social groups seem to have agendas and worldviews that cannot be captured by the … prototypes’ (Renn, 1992: 75). Grid/group analysis has also been criticised as too static as it cannot explain how risk perceptions of both individuals and organisations may change over time (Bellaby, 1990). Further, cultural theories are argued to have limited translation into empirical research (Fitchen et al, 1987; O’Riordan et al, 1997; Sjoberg, 1997). So, ‘While cultural theory has achieved prominence as a useful ideological weapon, the methodological rigour required for empirical cultural analysis has discouraged its practice’ (Rayner, 1992: 113).
Sociology and risk
In recent years there has been a great increase in sociological interest in risk. Like anthropologists, sociologists start from the premise that risk is socially constructed and collectively perceived (Gabe, 1995). However, in addition to examining cultural influences, sociologists also focus on the impact of material constraints and social interests on people’s experiences of risk.
Risk Society. Key social theorists on risk include Giddens (1991) and Beck (1992). Examining the nature of life in late modern society, Giddens (1991) has suggested that ‘to live in the universe of high modernity is to live in an environment of chance and risk’ (p.109). A key aspect of societal change for Giddens is the decline of trust in expert authority, where expert judgements are scrutinised, contested, accepted or rejected on the basis of lay people’s own assessments of risks. A central tenet of Beck’s (1992) thesis of ‘reflexive modernisation’ is that through the processes of modernisation, society has changed to become reflexive and intensely risk aware. For Beck, risk has become a means by which society deals with the hazards and insecurities that have been introduced by the processes of modernisation. He argues that these hazards are very different from those experienced by our predecessors ‘through the global nature of their threat … and through their modern causes’ (p.21). He particularly draws attention to the social and political dynamics of risk. Products once thought to be harmless turn out to be dangerous, and sources of wealth once admired emerge as unpredictable sources of danger. Risks develop political potential as they are seen as having social, economic and political consequences for which someone must take responsibility. The issue of knowledge of risk is an important aspect, as the invisibility of risks means that ordinary people are reliant on conflicting expert scientific sources for knowledge of hazards. This theory of risk society has been highly influential, and several social theorists have commented on its significance, as well as critiquing some of its basic premises (Wynne, 1996). It has also impacted upon wider policy debates about risk (Franklin, 1998).
Sociological risk research. Against this theoretical background, the need for sociological risk research has been highlighted (Douglas, 1985; Kronenfeld and Glick, 1991; Turner and Wynne, 1992). As Lupton (1993) has argued, ‘There needs to be a move away from viewing risk perception as a rational cognitive process … to more critical and theoretically informed investigations into the meaning of risk to individuals in contemporary society’ (p.433). Sociological research on risk perception and risk behaviour has used an interpretative methodological approach, examining how people interpret and respond to risk within the constraints of their daily lives. Key work includes perceptions of genetic risk (Parsons and Atkinson, 1992, 1993), and HIV/AIDS risk behaviour, which is seen as a ‘situated product’ (Bloor, 1993) of the immediate risk situation. The concept of ‘lived risk’ (Conners, 1992) has been developed, where risk is experienced, subjective and qualitative. Also important is sociological research on lay knowledge of health and science which has challenged the assumed superiority and objectivity of ‘expert’ risk knowledge, and pointed to ‘lay expertise’ (Kerr et al, 1998) in knowledge of risk (Irwin and Wynne, 1996; Williams and Popay, 1994).
Sociologists have also been concerned with social and political aspects of risk communication. Research has focused on the framing of risk debates by social institutions and groups such as the media (Kitzinger, 1993; Kitzinger and Reilly, 1997), and the complex technical and political choices involved in communicating uncertain scientific risk information (Nelkin, 1985). Discourses on risk are argued to serve ideological and political functions by blaming individuals or groups who may be seen as posing a risk to others (Lupton, 1993). Certain institutions may have particular power to define risk (e.g. science, government), and therefore assumptions of the desirability of risk communication fail to recognise the political dynamics of risk discourses which may be used to further the interests of powerful groups.
Having outlined the way risk has been conceptualised and studied across a range of social science disciplines, we turn now to briefly consider its use in the context of social welfare policy and practice.
Risk in social welfare policy and practice
We have already noted how recent writing on risk on social work practice has taken a sceptical turn, and consequently been marked by attention to previously taken-for-granted aspects of the ways risk assessment and management operate in social work. In particular, several writers have reflected on social welfare developments, arguing that concerns with risk assessment and risk management have replaced a focus on needs to become the driving force behind service provision (Burke, 1997; Kemshall et al, 1997; Parton, 1998; Parton et al, 1997; Waterson, 1999). While risk analysis is most prominent in the criminal justice and child protection fields, the mental health field, through preoccupations with notions of dangerousness, has also adopted risk terminology. The general argument of these writers is that as issues of rationing and accountability become more dominant, so do concerns with risk. However, in the context of social work, some writers have commented that technical terms such as ‘risk management’ simply reflect practices that social workers have always carried out, even if not formally described as such (Parsloe, 1999).
A key text on the concept of risk assessment in social work is Parsloe’s (1999) recent edited collection. Discourses around risk and risk assessment can be identified across a range of areas of social care and social work (Burke, 1997; Beaumont, 1999; Caddick and Watson, 1999; Stevenson, 1999; Waterson, 1999). Yet they are perhaps most prominently seen in the areas of child welfare (Parton, 1998; Sargent, 1999), and mental health work (Langan, 1999). In child welfare, risk is closely associated with the idea of significant harm, which underpins the Children Act (1989). The underlying policy principle is that risks to vulnerable service users are to be identified, measured and controlled. Social workers have increasingly become risk managers, spending much of their time ‘undertaking assessments of the risks to which children are exposed to enable them to plan services and make decisions about their future’ (Sargent, 1999: 179). In the field of mental health work, the concern is primarily with risk factors which allow prediction of whether an individuals poses a risk to others (e.g. violence) or risk to self (e.g. suicide). The policy and practice implications of this risk assessment approach are major as such judgements ‘frequently have profound and far-reaching implications for the individuals concerned, whether by curtailment of civil rights or a failure to protect individuals from their own actions’ (Langan, 1999: 156). Yet the use of the risk assessment approach in this context is problematic, as it is very difficult for mental health workers to accurately predict risk, and keep up-to-date with knowledge about risk factors. Further, an individualised perspective on risk factors fails to take into account the complex contextual nature of risk, shaped by social, economic and political factors (Langan, 1999).
Blame and accountability are also key risk-related concepts becoming increasingly prevalent in social work discourses. Parton (1998) argues that concepts of risk have shaped the way social workers organize themselves and are organized by others. Different organizational systems of accountability and blame have emerged. Within these systems, social workers are increasingly made responsible for risk management, faced by the threat of blame when things go wrong, when vulnerable individuals or wider society are conceived to have been placed at risk from people defined as dangerous.
However, the focus on risk in social work, with its emphasis on accurate risk assessment and prediction, has been criticised by some. Parton (1998) has argued that the implications of ‘calculability and objectivity’ inherent to the risk concept are problematic in the context of social work, where risk is ‘inherently contingent and open to differing and sometimes conflicting interpretations’ (Parton, 1998: 6). Therefore, greater focus on notions of uncertainty and ambiguity is called for in social welfare work, where much experience is not characterized by scientific calculations of risk, but imbued with intuition and uncertainty:
by their very nature, notions of risk are very uncertain and imperfect. Given the mobile character of the social world and the mutable and frequently controversial nature of abstract systems of knowledge, most forms of risk assessment contain numerous imponderables. Uncertainty and controversy are built into the knowledge, policies and practices concerned with risk
(Kemshall et al, 1997 : 228)
Implications of risk debates for social work research
So, drawing on these wider risk debates within the social sciences and increased concern with risk within social welfare policy and practice, what are some implications for social work research?
The key themes emerging from these risk debates which may impact how we think about social work research can be summarised as follows:
Figure 1: Key themes from risk literature
Epistemology of risk
- Risk as objective and quantifiable, existing independently of our perception, or
- Risk as subjective and qualitative, experienced in specific social contexts
Key risk terms/concepts
- Risk, hazard, danger, harm, safety, vulnerability, dangerousness, blame, accountability
Methodology and methods for risk research
- Quantitative and objective (control, rigour, precision), using structured scientific methods, e.g. questionnaires, psychometric instruments, or
- Qualitative and subjective (context, process, meaning), using ‘unstructured’ methods of sociology/social anthropology, e.g. depth interviews
Key critical themes
- ‘Lay’ vs ‘expert’ knowledge of risk: rationality of lay responses to risk; uncertainty and social basis of expert knowledge
- Social dimensions of risk: risk as, at least in part, socially-constructed and collectively perceived
- Political contexts of risk: power and agendas of risk communicators
- Trust in/credibility of ‘experts’ in a risk society
- Risk regulation: responsibility of individuals or institutions?
- ‘Ethics’ of risk: human rights to information and protection from harm; blaming and ‘scapegoating’ within risk situations
Conceptual and methodological debates about risk within psychology, cultural theory and sociology cause us to re-evaluate and problematise the use of the term risk within social work policy and practice. They also pose questions for social work research. Underlying epistemological debates about risk draw attention to the need to consider risk as both an ‘objective’ phenomenon, with real material consequences, yet also one that carries complex social and political dimensions. Conceptual discussions of risk highlight the need to problematise the risk terminology used within social work, and consider the close inter-relationships between an array of risk-related terms. Social science risk research poses questions about the methodologies and methods used within social work research, which may be rooted in assumptions about risk as objective and quantifiable, or subjective and social. Risk debates also raise a host of critical issues which may impact on social work research, including issues around ‘lay’ and ‘expert’ knowledge of risk, the social and political contexts of risk, matters around trust in, and credibility of, ‘experts’ in a risk society, concerns around responsibility for the regulation of risk, and wider ethical dimensions of risk.
More specifically, debate regarding risk and society develops the agenda for risk and social work research in several directions. First, it pushes us to think about research ethics in a risk context. Second, the apparent decline of trust in ‘experts’ suggests that we should look at how social work practitioners actually deal with risk. Insofar as it is the case that society is becoming intensely risk-aware, there needs to be research on how service users deal with risk, in much the same way as the work on genetic risk and HIV/AIDS cited above. Researchers are members of the risk society and not outside observers. Their own expertise is just as much on the line as that of practitioners. Conceptual, theoretical and empirical work within the social sciences has provided a rich source of critical debate about risk. While social work writing on risk has generally failed to draw on this, a minority of key writers such as Parton (e.g. 1998) have made the cross-over from general debates about risk society to social work ideas of risk. Drawing attention to the centrality of uncertainty to considerations of risk within social work, Parton points to a key consequence, in that service users, researchers and practitioners share aspects of this inherent uncertainty, which shapes and partly justifies collaborative and user led research.
Having considered risk debates within the social sciences and social work policy and practice, and drawn out some tentative general ideas about the way these impact on social work research, we now turn to consider in more detail how a wider awareness of risk may enrich the understanding and conduct of social work research. We will focus in particular on four areas which to date have remained unexplored in social work writings on risk: research ethics, the political contexts of research, the uses of research, and ‘risk talk’. In each of these areas, we will draw on specific examples of social work research which illustrate these ideas.
Risk and Research: Research ethics
Discussions of research ethics in social work have been seriously underdeveloped. For example, while we are reluctant to suggest additional layering of decision-making mechanisms, we believe that decision-making on ethical aspects of social work research has been skewed and narrowed by the assumption that local and multi-site medical ethics committees are the appropriate location for regulating ethical aspects of such research.
Questions of research ethics frequently turn on issues of risk, consent and trust. A research study most discussed for its ethical implications is Laud Humphreys’ Tearoom Trade (Humphreys, 1975). Humphreys adopted a covert ‘watchqueen’ role for his participant study of (then illegal) male homosexual activity. Some men knew of his research and others did not. He subsequently illegally obtained the addresses of men in the study and interviewed them, unrecognised, regarding a supposed ‘health survey’ (Humphreys, 1975).
Under the general rubric of human freedom, Warwick (1982) outlines the potential risks and benefits to human freedom of conducting research. Potential benefits include opportunity for self-expression through the findings, the satisfaction of sharing important things with someone else, the intellectual satisfaction of participating in research, and the possible personal benefits from talking about a problem. Potential risks include deception, the invasion of privacy, and the misuse of information. Warwick wonders whether we need to question liberal assumptions regarding the value of information. He asks, ‘Does research reinforce the tendency of individuals to be wary and to live for the record?’ (Warwick, 1982: 52). He suggests that research consumes resources of money, personnel, goodwill and tolerance, all of which are finite. While there is an obligation to study controversial topics, that does not include the right to deceive, exploit or manipulate.
The ethics of risk and deception, with the associated values of trust, are especially prominent in qualitative research. Given that much social work research has a broadly evaluative purpose, withholding the nature of the inquiry from participants is ‘a particularly dangerous example of covert social research, because the findings are not being used merely to illuminate our knowledge of the social world, but potentially to change it’ (Finch, 1986: 203). Managing such risk aspects of research ethics is especially uncertain in circumstances where some participants may be less sophisticated than others in raising issues.
Do researchers have an ethical responsibility to serve in a dual role: first, as researchers with a project aimed at satisfying their research purposes, and second, as advocates for the practitioners, raising questions that the researchers know should be raised in order for practitioners to make a competent assessment of the risks? (Eisner, 1991: 217)
We think the answer to Eisner’s question is ‘yes’.
Political context of research
It may appear puzzling to social work researchers in Britain and the west that researchers in most countries that previously made up the USSR, and also in most countries of the Arab world, rely almost entirely on survey methods of research. It is only when survey methods are seen through the prism of western political pluralism that the problem comes into focus. Take, for instance, Hakim’s (1988) argument that the survey is the most democratic form of research design, in that methods and procedures can be made visible and accessible to non-specialists and specialists working in other disciplines. This, she concludes, makes the survey more transparent and accountable than many other methods (Hakim, 1988). Whether or not her case holds water in every instance, there do appear to be political assumptions embedded in the apparently technical decisions associated with survey research.
Simple random sampling, for example, carries the assumption that the views of one participant have equal weight to the views of all other sampled participants. This assumption may well be challenged in traditionally hierarchical societies, both by those who hold power and those from power is excluded. Related to this, ‘conventional’ (i.e. through western eyes) methods of survey analysis are assumed by researchers and their audience to be relatively neutral and value free – ie technical rather than political matters. Survey researchers, it is assumed, will almost routinely uncover evidence of variance and will produce tables which differentiate those who respond according to this or that factor. It is not far from the truth to say, however, that far from being neutral such methods of sampling and analysis assume a pluralist society where, within certain normative boundaries, differences of opinion are accepted as part and parcel of ‘normal’ society (Shaw and Al-Awwad, 1994).
Why then the preponderance of survey designs? Hakim’s transparency holds the key. The dominance of survey methods has potent political attractions that address issues of risk. Researchers who work with large samples, for example, cannot easily be accused of selective coverage. Data analysis that rests on standardised response formats can less easily be suspected of presenting doubtful or arguable interpretations. Hence, in the context of this Arab world research, surveys provided a politically risk-free method of mounting a limited critical position, and enabled sociology to survive in a one-party state (1).
More pointed risks face researchers whose inquiries directly challenge the position of those who hold power. Gomez-Cespedes describes her research on organised crime in a country where law enforcement agencies and intelligence services are part of the problem and where ‘promoting academic research on criminological and policing issues becomes a threat to deep and powerful interests’ (1999: 166). Her fascinating account of the combination of unpredictable factors that gave her access to the heart of her country’s federal government agency defeats simple conclusions, but underscores the complex interplay of trust, risk and serendipity that shape research in risky political contexts.
The uses of research
The political context of research and the potential threats involved are often discussed in relation to the form of research reporting and the uses to which findings are put. Writing research, and the relinquishing of research into the policy and public domain both carry profound risk implications. This is as it should be, and we should not be precious or purist regarding the work we undertake. One of us was involved in a research project funded by the former Scottish Office and the European Commission. This research intended to provide estimates of the numbers of people sleeping rough across Scotland, and the feasibility of developing and utilising a prevalence estimating methodology usually known as mark-recapture (or contact-recontact) research (see Shaw, Bloor, Cormack and Williamson, 1996 for a description of the methodology and its potential application to the field of homelessness research). The funding for the research followed strong official support for the work.
The research failed to deliver the estimates of prevalence or the hope that the methodology could be used more generally as a policy tool in the UK or within collaborating EU member states. However, the method had previously been used effectively for the estimation of population prevalence in other fields, and has been used subsequently. The conclusion of the fieldwork and analysis posed several sharp risk-related problems. How should the research be presented? The funders pressed the research team to dredge whatever useable information possible from the project. Should the research team resist or accept the risk that the research will be over-interpreted? What were the likely consequences for government confidence in research, especially when that research pursues partly-tried and innovative methodology? Are researchers in any way responsible for resolving difficulties surrounding stewardship of public monies? How should the work be reported?
In this instance there were inevitable differences of view in the research team. The research was made available in the form of a technical report (Shaw, Bloor and Roberts, 1996) with no policy recommendations. The agencies and organisations supporting the project and providing access were probably less likely to collaborate in the future with a methodology of this kind, and the government department less likely to risk funding innovative research. Yet perhaps government departments take a swings and roundabouts view of research funding and are sanguine enough to accept that they win some and lose some. In any event, the publication of the technical report was lost in the simultaneous announcement by a government minister that funding was to be made available for a rough sleepers initiative in Glasgow, paralleling existing initiatives in England.
There is at least one lesson to be drawn from this experience. The processes and risk dimensions of research utilisation are much discussed but too little understood. Weiss’ work in the USA in the 1970s radically altered the understanding of how policy research and larger programme evaluations are likely to enter the policy stream (Weiss, 1980, 1988; cf Finch, 1986, Bulmer, 1982; Shaw, 1999). Weiss’ work remains largely unfamiliar to the social work community, and it may be timely for corresponding empirical work to be undertaken. However, greater understanding of the utilisation process will not remove the risk and uncertainty.
The ethical and political dilemmas about how research material is used are sometimes sharper in qualitative research than in quantitative work. For example, a risk of betrayal may arise partly from the greater closeness and consequent trust that may develop between researcher and participant. In quantitative research the greater distancing may make these issues less agonising. The risk of betrayal is increased because of the characteristic use of smaller samples, and the emphasis on the details of how people live their lives.
Finch describes from her playgroups evaluative research her ‘sense that I could potentially betray my informants as a group, not as individuals’ (Finch, 1986: 207). ‘Where qualitative research is targeted upon social policy issues, there is the special dilemma that findings could be used to worsen the situation of the target population in some way’ (Finch, 1985: 117). Finch’s particular interest was in what self-help playgroup provision would mean for working class women living in economically deprived areas. Over a three-year period, through observation and semi-structured interviewing, she was able to document the character of self-help playgroups in such areas:
I uncovered situations where practice diverged wildly from bourgeois standards of child care and education which most policy makers and academics would take as the norm, and at times were downright dangerous. (1985: 117)
Finch was worried that the publication of her work would risk further reinforcing ‘those assumptions deeply embedded in our culture and political life that working class women (especially the urban poor) are inadequate mothers’ (p.117). Those who had welcomed her for three years would thus be betrayed. She had to work through these problems by being reflexive about the process of her research. Had she been guilty of taking a middle class norm and imposing it on these groups? Yet that norm was the one to which the women who ran the groups aspired. It was the participants’ model and not simply hers. She eventually developed reasoning which avoided the ‘deficit’ model of explanation, and argued that to view working class mothers as incompetent is improper and naïve. Uncertainty was central to her partial resolution of her role in the research and the risks inherent to the use of its findings. Finch acknowledged her uncertainty as to whether she has fully resolved the issues, and accepted that,
To argue like this is to take a frankly moral stance, far removed from the model of the objective social scientist… It seems to me that qualitative research on social policy issues will lead inevitably to explicit moral stances of that sort, and that it can never simply provide the ‘facts’. (1985 : 119-20)
Conceptual work within the social sciences, in particular cultural and social theories of risk which highlight differing interpretations and perspectives on risk, and point to the wider decline of public trust in ‘experts’, has bearing upon our understanding of the way risk is talked about within social work. They point to the value of understanding how social workers assess, and claim as theirs, the attribution of risk among service users. White concludes that recent social work analysis of how an emphasis on risk has contributed to the withering away of the more preventive and supportive aspects of social work
neglect(s) the ways in which notions of risk, dangerousness, deviance and normality are reproduced and instantiated in the cut and thrust of everyday talk among social workers.
(White, 1999: 87)
Other research has elicited recurring instances of ways in which practice regarded by social workers as having gone less than well is likely to trigger deep uncertainties in the mind of the reflective practitioner (Shaw and Shaw, 1997). Am I responsible? Should I take the blame? Could it have been avoided? The dilemma and sense of risk are unforgivingly sharp. There are two realities that need to be brought into conjunction.
First, social workers exhibit a widespread sense that outcomes, both good and bad, are beyond prediction and will never cease to jump out and surprise the practitioner. Social workers in this research were characterised by this uncertainty. Second, social workers display a hesitation about claiming credit for ‘good’ work or blame for ‘bad’ work. The two realities – predicting and accountability – are linked. In the words of one probation officer:
I think, bearing in mind the fact that you can’t predict what judges are going to do, you can’t predict what magistrates are going to do, in a sense you can’t take the blame for a failure. If you can’t take the blame if it doesn’t work, how can you take the credit if it does?
Social workers widely share the conviction that the evidence facing them in day to day practice will always prove more or less ambiguous and complex, and open to competing interpretations:
I personally don’t feel a hundred per cent certain about anything I do. I think it’s so difficult when you’re working with individuals.
One practitioner employed an apt simile to suggest:
It’s like snap photographs – one individual photograph isn’t going to tell you very much. Over a period of time, that’s how we see things, but that’s not easy.
Social worker accounts of work that is going well or not may be regarded as simultaneous attempts to explain a problem, and secure a professional safety net. Accounts may secure colleague and management approval for limited achievement in the past and limited ambitions for the future (cf Bull and Shaw, 1992). Social workers offer both retrospective and prospective rationalities for limited and imprecise effects from social work intervention, and for an incremental style of work described by one of our respondents as ‘chipping away’.
In the first part of the paper we reviewed discussions regarding the conceptualisation of risk and developments in theories of risk, and made cautious suggestions of links and gaps between this work and recent social work writing on risk. We also drew out some general implications from these risk debates for social work research. In the second part of the paper we illustrated how a wider research awareness of risk – as central to the social, political, substantive and methodological contexts of disciplined inquiry – provides a strengthened purchase on a critique of and commitment to the social purpose of social work. This provisional analysis of the implications of risk debates for social work research points to the ways in which ideas of risk give shape and texture to research ethics, awareness of the political context for research, the uses of research, and the street-level fashioning of risk strategies in social work teams. Attention to such social and political dimensions of risk can enrich our understanding of the complex processes of social work research within a risk society.
(1) The problem is more complex than we have discussed here. For instance, in Muslim societies the great majority of researchers are men, and where face to face interviewing of women, more participant methods of inquiry, and even tape recording of interviews, pose significant problems. Survey methods have the attraction of data collection at a distance.
Alaszewski, A. (1998) ‘Risk in Modern Society’, in A. Alaszewski, L. Harrison and Manthorpe, J. (eds), Risk, Health and Welfare, Buckingham: Open University Press.
Beaumont, B. (1999) ‘Assessing Risk in Work with Offenders’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Beck, U. (1992) Risk Society, London: Sage.
Bellaby, P. (1990) ‘To Risk or Not To Risk: Uses and Limitations of Mary Douglas on Risk Acceptability for Understanding Health and Safety at Work and Road Accidents’, Sociological Review, 38: 465-483.
Bloor, M., Barnard, M., Finlay, A. and McKegany, N. (1993) ‘HIV-Related Risk Practice Among Glasgow Male Prostitutes: Reframing Concepts of Risk Behaviour’, Medical Anthropology Quarterly, 7, 1-19.
Brearley, C.P. (1982) Risk and Social Work, London: Routledge and Kegan Paul.
Broadbent, D.E. (1985) ‘The Psychology of Risk’, in M.G. Cooper (ed), Risk: Man Made Hazards to Man, Oxford: Clarendon.
Bull, R. and Shaw, I. (1992) ‘Constructing causal accounts in social work’ Sociology 26(4): 635-649.
Bulmer, M. (1982) The Uses of Social Research: Social Investigation in Public Policy-Making London: Allen and Unwin.
Burke, P. (1997) ‘Risk and Supervision: Social Work Responses to Referred User Problems’, British Journal of Social Work, 27(1): 115-129.
Burke P. (1999) ‘Social services staff: risks they face and their dangerousness to others’ in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Caddick, B. And Watson, D. (1999) ‘Rehabilitation and the Distribution of Risk’, In P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
CCETSW (1995) Assuring Quality in the Diploma in Social Work: Rule and Requirement for the DipSW, London: CCETSW.
CCETSW (1997) Assuring Quality: Requirements for the Post Qualifying and Advanced Awards in Social Work, London: CCETSW.
Conners, M.M. (1992) ‘Risk Perceptions, Risk Taking and Risk Management Among Intravenous Drug Users: Implications for AIDS Prevention’, Social Science and Medicine, 34(6): 591-601.
Department of Health (1989) Children Act, London: HMSO.
Douglas, M. (1966) Purity and Danger: An Analysis of Concepts of Pollution and Taboo, London: Routledge and Kegan Paul.
Douglas, M. (1985) Risk Acceptability According to the Social Sciences, London, Routledge.
Douglas, M. (1990) ‘Risk as a Forensic Resource’, Daedalus 119(4), Fall: 1-16.
Douglas, M. (1992) Risk and Blame: Essays in Cultural Theory, London: Routledge.
Douglas, M. and Wildavsky, A. (1982) Risk and Culture, London: University of California Press.
Eisner, E. (1991) The Enlightened Eye: Qualitative Inquiry and the Advancement of Educational Practice, New York: Macmillan.
Finch, J. (1985) ‘Social policy and education: problems and possibilities of using qualitative research’, in Burgess, R. (ed) Issues in Educational Research: Qualitative Methods, London: Falmer Press.
Finch, J. (1986) Research and Policy: the Uses of Qualitative Methods in Social and Educational Research, London: Falmer Press.
Fitchen, J.M., Heath, J.S. and Fessenden-Raden, J. (1987) ‘Risk Perception in Community Context: A Case Study’, in B.B. Johnson and V.T. Covello (eds), The Social and Cultural Construction of Risk, Boston: Reidel.
Franklin, J. (ed) (1998) The Politics of Risk Society, Cambridge: Polity Press.
Fraser, M.W., Richman, J. and Galinsky, M. (1999) ‘Risk, protection and resilience: towards a conceptual framework for social work practice’ Social Work Research, 23(3): 131-143.
Gabe, J. (ed) (1995) Medicine, Health and Risk, London: Blackwell.
Giddens, A. (1991), Modernity and Self Identity, Cambridge: Polity Press.
Gomez-Cespedes, A. (1999) ‘Organised crime in Mexico’ in Brookman, F., Noaks, L. and Wincup, E. (eds) Qualitative Research in Criminology, Aldershot: Ashgate.
Hakim, C. (1988) Research Design, London: Unwin Hyman.
Hansson, S.E. (1989) ‘Dimensions of Risk’, Risk Analysis, 9(1), 107-112.
Hayes, M.V. (1992), ‘On the Epistemology of Risk: Language, Logic and Social Science’, Social Science and Medicine, 35(4): 401-407.
Heimer, C.A. (1988) ‘Social Structure, Psychology and the Estimation of Risk’, Annual Review of Sociology, 14: 491-519.
Humphreys, L. (1975) Tearoom Trade. Chicago: Aldine.
Irwin, A. and Wynne, B. (1996) ‘Introduction’, in A. Irwin and B. Wynne, (eds) Misunderstanding Science? The Public Reconstruction of Science and Technology, Cambridge: Cambridge University Press.
Kemshall, H. and Pritchard, J. (eds) (1996) Good Practice in Risk Assessment, London: Jessica Kingsley.
Kemshall, H. (1998) Risk in Probation Practice. Aldershot: Ashgate.
Kemshall, H., Parton, N., Walsh, M. and Waterson, J. (1997) ‘Concepts of Risk in Relation to Organizational Structure and Functioning Within the Personal Social Services and Probation’, Social Policy and Administration, 31(3): 213-232.
Kerr, A., Cunningham-Burley, S. and Amos, A. (1998) ‘The New Genetics and Health: Mobilising Lay Expertise’, Public Understanding of Science, 7(1): 41-60.
Kitzinger, J. (1993) ‘Understanding AIDS: Researching Audience Perceptions of Acquired Immune Deficiency Syndrome’, In J. Eldridge (ed), Getting the Message, London: Routledge.
Kitzinger, J. and Reilly, J. (1997) ‘The Rise and Fall of Risk Reporting: Media Coverage of Human Genetics Research, ‘False Memory Syndrome’ and “Mad Cow Disease”’, The European Journal of Communication, 12(3), 319-350.
Kronenfeld, J.J. and Glick, D.C. (1991) ‘Perceptions of Risk: Its Applicability in Medical Sociological Research’, Research in the Sociology of Health Care, 9: 307-334.
Langan, J. (1999) ‘Assessing Risk in Mental Health’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Lupton, D. (1993), ‘Risk as a Moral Danger: The Social and Political Functions of Risk Discourse in Public Health’, International Journal of Health Services, 23: 425-435.
MacDonald, K. and MacDonald, G. (1999) ‘Perceptions of Risk’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Nelkin, D. (1985) The Language of Risk, Beverley Hills: Sage.
O’Riordan, T., Marris, C. and Langford, I. (1997), ‘Images of Science Underlying Public Perceptions of Risk’, Royal Society Paper, London, 1997.
Parsloe, P. (1999) ‘Introduction’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Parsons, E. and Atkinson, P. (1992) ‘Lay Constructions of Genetic Risk’, Sociology of Health and Illness, 14: 437-455.
Parsons, E. and Atkinson, P. (1993) ‘Genetic Risk and Reproduction’, Sociological Review, 41: 679-706.
Parton, N. (1998) ‘Risk, advanced liberalism and child welfare: the need to rediscover uncertainty and ambiguity’, British Journal of Social Work, 28 (1): 5-27.
Parton, N, Thorpe, D. and Wattam, C. (1997) Child Protection, Risk and the Moral Order, London: Macmillan.
Rayner, S. (1992) ‘Cultural Theory and Risk Analysis’, in S. Krimsky and D. Golding (eds), Social Theories of Risk, Westport, CT: Praeger Publishing.
Reamer, F.G. (1998), The Evolution of Social Work Ethics’, Social Work, 43(6): 488-500.
Renn, O. (1992) ‘Concepts of Risk: A Classification’, in S. Krimsky and D. Golding (eds), Social Theories of Risk, Westport, CT: Praeger Publishing.
Sargent, K. (1999) ‘Assessing Risks for Children’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Shaw, I. (1999) Qualitative Evaluation, London: Sage.
Shaw, I. and Al-Awwad, K. (1994) ‘Culture and the indigenisation of quality in third world social research’ Journal of Social Development in Africa, 9(1): 59-71.
Shaw, I. and Shaw, A. (1997) ‘Game plans, buzzes and sheer luck: doing well in social work’ Social Work Research, 21(2): 69-79.
Shaw, I., Bloor, M. and Roberts, S. (1996) Without Shelter: Estimating Rooflessness in Scotland, Edinburgh: Central Research Unit, Scottish Office.
Shaw, I., Bloor, M., Cormack, R and Williamson, H. (1996) ‘Estimating the Prevalence of Hard to Reach Populations: Illustration of Mark-Recapture Methods in the Study of Homelessness’, Social Policy and Administration, 30(1): 69-85.
Sjoberg, L. (1997) ‘Explaining Risk Perception: An Empirical Evaluation of Cultural Theory’, Risk Decision and Policy, 2(2): 113-130.
Slovic, P., Fischoff, B. and Lichtenstein, S. (1980) ‘Facts and Fears: Understanding Perceived Risk’, in R. Schwine and W. Albers (eds) Societal Risk Assessment, New York: Plenum Press.
Stacey, J. (1988) ‘Can there be a feminist ethnography?’, Women’s Studies International Forum, 11(1): 21-27.
Stevenson, O. (1999) ‘Old People at Risk’, in P. Parsloe (ed), Risk Assessment in Social Care and Social Work, London: Jessica Kingsley.
Turner, G. and Wynne, B. (1992) ‘Risk Communication’, in J.Durant (ed), Biotechnology in Public: A Review of Recent Research, London: Science Museum for the European Federation of Biotechnology.
Tversky, A. and Kahneman, D. (1982) ‘Judgement Under Uncertainty: Heuristics and Biases’, in D. Kahneman, P. Slovic and A. Tversky, Judgement Under Uncertainty: Heuristics and Biases, New York: Cambridge University Press.
Warner, F (1992) ‘Introduction’, in Risk: Analysis, Perception and Management, Report of a Royal Society Study Group, London: Royal Society.
Warwick, D. (1982) ‘Tearoom Trade: means and ends in social research’, in Bulmer, M. (ed) Social Research Ethics, London: Macmillan.
Waterson, J. (1999) ‘Redefining community care social work: needs or risks led?’ Health and Social Care in the Community. 7(4): 276-279.
Weiss, C. (1980) ‘Knowledge creep and decision accretion’, Knowledge, Creation, Diffusion, Utilisation, 1(3): 381-404.
Weiss, C. (1988) ‘If programme decisions hinged only on information’ Evaluation Practice, 9(1): 15-25.
Wharton, F. (1992) ‘Risk Management: Basic Concepts and General Principles’, in J. Ansell and F. Wharton (eds), Risk: Analysis, Assessment and Management, New York: John Wiley.
White, S. (1999) ‘Examining the artfulness of ‘risk talk’ in Jokinen, A., Juhila, K. and Pösö, T. (eds) Constructing Social Work Practices, Aldershot: Ashgate.
Williams, G. and Popay, J. (1994) ‘Lay Knowledge and the Privilege of Experience’, in D. Kelleher and G. Williams (eds) Challenging Medicine, London: Routledge.
Wynne, B. (1996) ‘May the Sheep Safely Graze: A Reflexive View of the Expert-Lay Knowledge Divide’, in S. Lash,, B. Szerszynski and B. Wynne (eds) Risk, Environment and Modernity, London: Sage