Responses to complexity: similarities in knowledge generation and governing paradigm

A growing number of scholars have criticised the negative impact of the neoliberal agenda that has become a new orthodoxy in Europe and elsewhere in the western industrialised world. Key issues include the shift of risk and responsibility from the state to the individual (Hacker, Reference Hacker2008), the restricted understanding of citizenship (Taylor-Gooby, Reference Taylor-Gooby2009), the simplistic understanding of action caused by a decontextualised rational actor model (Taylor-Gooby, Reference Taylor-Gooby2009), and the reliance on quantitative rather than qualitative evidence (Rudman, Reference Rudman2006). The common, sometimes generalised, critique of the long-term change towards market liberalism, neo-liberalism and economic rationalism might give the impression that political ideology, style of governing and forms of regulation fit together as a coherent policy regime that has particular understandings about the production of authoritative knowledge. By contrast, we suggest that it is more appropriate to see these components as loosely coupled, and as one possible response to the challenges of governing increasingly complex societies (Foucault, Reference Foucault1978).

In his historical analysis, Foucault identified a shift from the dominance of punishment in the sovereign society to surveillance in the discipline society (Foucault, Reference Foucault1977) and to governmentality in the modern state (Foucault, Reference Foucault1978). With growing societal complexity, direct forms of control were increasingly supplanted by more indirect forms of governing. He argued that governments produce conditions for the self-governing of individuals (‘technologies of the self’) supporting a discourse of autonomous individual subjects independently improving their lives and wellbeing, while using generalised knowledge and population data to manipulate societal conditions. These views are supported by Hacking's (Reference Hacking1990) analysis of the development of statistics and how these are linked to new forms of governing societies. It is crucial for this new kind of probabilistic knowledge that individuals are no longer approached as a whole but as bearers of indicators or as part of artificially constructed populations addressed by specific interventions. Thus, this new form of knowledge keeps complex societies manageable.

There is a second advantage in the generation of objective knowledge (Porter, Reference Porter1995). Probabilistic methods not only allow the transformation of unmanageable contingent futures into complex and manageable ones, but the alleged objectivity enables management of the growing diversity of values and world views of modern highly differentiated and multicultural societies. Unbiased knowledge, however unlikely, is important for public servants and politics because it supports the justification of political decisions and provides ‘proof’ of the quality of service.

We suggest that both the reliance on demographic or standardised survey data, and the neoliberal agenda, share the advantage of substantially reducing the complexity of governing increasingly complex societies (Luhmann, Reference Luhmann1995). Large surveys have proven useful in identifying general societal trends and changes in behaviour and attitudes. They can therefore serve as a powerful tool for critique and defence of policy paradigms because they carry a certain degree of ‘neutrality’ and ‘objectivity’ (see, for example, Moran, Reference Moran2007; Horn et al., Reference Horn, Scutella and Wilkins2011; Productivity Commission, 2012). The focus on generalised patterns, associations and correlations is useful to identify underlying patterns and forces which influence social behaviour. However, the normative model of rational actions may be too narrow to capture different rationalities, such as those that shape different groups’ decisions about paid work (Ranzijn et al., Reference Ranzijn, Carson and Winefield2004; Rudman, Reference Rudman2006; Bowman et al., Reference Bowman, Bodsworth and Zinn2012). Survey research is ill-equipped to identify and explain complex or new issues because predefined questions prevent the formulation of competing interpretations that could challenge the underlying model of the subject or social processes. At the same time, questions that do not relate to the realities experienced by participants reinforce disadvantage and marginalisation. We will argue in the following section that standardised surveys, in particular, tend to reproduce the experience of disempowerment of participants through reinforcing dominant perspectives.

Responding to questions: a case study

To illustrate our argument, we draw on a small study ‘Qualitative exploration of the perception and behaviours of the HILDA survey participants’ (Boehm, Reference Boehm2012) that aimed to explore the sense-making processes in survey research. The study examined fifty-three questions relating to housing from the Household Income and Labour Dynamics in Australia (HILDA) survey wave 10 (Melbourne Institute of Applied Economic and Social Research, 2010).Footnote ² The HILDA survey is a nationally representative longitudinal survey that is extremely influential, as many studies draw on the survey data to inform policy development in Australia. Indeed, a recent publication states, ‘The HILDA study has regularly informed Productivity Commission work since the release of the first wave of data in 2002’ (Melbourne Institute of Applied Economic and Social Research, 2010; HILDA, 2011: 2). For this reason, it is an interesting example of the role of statistical survey data in policy development.

The study entailed interviews with a small number of older men and women in Broadmeadows, a disadvantaged suburb on the fringe of Melbourne. Older long-term residents were recruited purposively (three women and three men; half of them were renters and half home owners. All were retired and had lived in the area for at least fifteen years). Participants were asked to complete the fifty-three survey questions and to reflect on their feelings about and reactions to the questions. The study focussed on the survey encounter rather than the collection of answers to the questions. To set their reactions in some context, participants were asked open-ended biographical questions about their housing trajectories. Subsequently, a focus group was conducted with four of the six participants to provide further insight into the survey encounter. Participants were asked to critically reflect on the increasingly familiar experience of survey completion. Asking the participants to critically reflect on these questions may have reinscribed their sense of marginalisation, the questions did not apply to them. Yet the participants resisted symbolic domination by affirming their identities as renters and as proud working-class residents of Broadmeadows. While the small exploratory study has limitations, it provides useful insight into participants’ experience of the generation of statistical data, a key element of the production of statistical knowledge that is often overlooked.

When asked to consider specific questions from the HILDA survey, only one participant was openly critical of the survey questions. Most participants responded with silence, resignation, cooperation and, at times, sarcasm, rather than direct criticism. A common pattern was that apparently straightforward questions often needed participants to refer to sources other than their personal knowledge or memories. For example, questions relating to land value required reference to official documents such as rates notices or Office of Housing land value assessments. It is important to note here that the HILDA questionnaire is designed to enable irrelevant questions to be skipped. By contrast, in this study, all participants considered all fifty-three questions relating to housing, whether or not they were relevant to their situations. Nevertheless, the official nature of the information caused anxiety. Where participants were unable provide information there was a sense of inadequacy. For example, a participant was unable to answer a question about rent.

This man was unable to provide details because as a public tenant his rent is calculated as a percentage of his income support payment. This situation is at odds with the questionnaire's assumption that participants would know how much rent they paid. The questions are asked in a social policy context where public housing is marginalised and residual, thus the experience of responding to such questions reinforced the negative framing of the participants’ situation.

This study highlights the interdependence between administrative rules, regulations, laws and individual circumstances. The participants’ understanding of relevant financial information relating to housing was shaped by their experience of eligibility criteria for public housing and administrative rules, such as those used to calculate rent on the basis of income or family composition. As older people, the participants retained aspects of a working-class identity, yet they explained their housing trajectories in terms that were infused with a neo-liberal ideology. They did not call into question the normative ideological framing of the survey questions, nor did their critique relate to broader conceptualisation of housing policies. For these participants, the ‘Housing Commission’ embodied nearly all aspects of the fragmented housing policies delivered by the local Council, the State and Commonwealth governments. ‘Housing Commission’ is a slightly derogatory and outdated term for public housing. The Housing Commission of Victoria was established in the 1930s and was responsible for the demolition of slums and the construction of large tower blocks in the 1960s, which are often referred to as Commission flats. It was also responsible for the development of cheap housing for factory workers in suburbs like Broadmeadows (Faulds, Reference Faulds2002; Lemon, Reference Lemon1982).

The study found that the survey questions failed to ask relevant questions that could account for the complexity that was revealed through the biographic interviews. For example, the unbalanced power relationship between the public housing authorities and tenants, and how regulations and policies construct identities, prevent participation and shut-down the formulation of critique. The narrow framing of the survey questions prevented participants from providing information they judged significant and important, and thus reinforced the marginalisation of their experience and perspectives. Housing surveys form part of a long-standing dominant administrative relationship between public housing tenants and the public authorities, yet this relationship is obscured by the decontextualised creation of data. For these individuals, criticism of the Housing Commission was a substitute for a critical conceptualisation of housing policy. They focussed on recounting negative personal encounters with Housing Commission staff, for example one man referred to staff talking to him as if he were ‘absolute rubbish’. These narratives filtered out broader structural critiques. For example, a participant stated:

This expressed his understanding of the way in which the inspectors perceived him. He neutralised it and expressed resistance on an individual level to the inspector. The last words, ‘but anyway’, are an evasive conclusion, as if he was referring to something without remedy, except on a personal level.

The participants in this study were resigned to the impossibility of political representation. Yet they implicitly understood that the method of data collection reflects and reinforces the power relationship with public authorities. Participants repeatedly stated that political consultations or surveys did not provide the expected outcomes. For these individuals, the completion of surveys took place in the context of long experience that their opinions would not be taken into account and their expectations would be disappointed. Indeed, in the selected HILDA questions there was no opportunity for respondents to provide their opinions. The repetition of the expression ‘they don't listen’ and of the narrative of frustrating political consultation reflected their sense of disenfranchisement. Furthermore, the conflation of political consultation with social research suggests that for these individuals knowledge generation is political. Yet as long-term residents of a disadvantaged area, they were acutely aware of their lack of power. As one interviewee observed:

The awareness of the social inequalities and material conditions that determined political participation underlined the interviewees’ claim that their voices remained unheard. This may explain their attitudes toward the survey questions. Indeed, they anticipated the tenuous and discredited nature of their claims.

Administrative practices that are hard to decipher, the retention of personal information, the unequal power relationship and the participants’ lack of cultural capital (to manipulate administrative rules and to dare to speak during formal meetings) all work together to silence their claims. As a participant observed, this is an impediment to formulating alternatives:

Government administration has a symbolic monopoly over what is feasible and practical, thus policy discourses secure a monopoly over realist and pragmatist perspectives, obstructing other social discourses (Bourdieu, Reference Bourdieu1999). In turn, this monopoly is associated with government's central role as collector and creator of information (Hood, Reference Hood, Smelser, Wright and Baltes2001). By reasserting the social order and its value (Thévenot, Reference Thévenot2011), the survey re-establishes the limit of what could be possibly thought and claimed (Bourdieu, Reference Bourdieu1999). It feeds an understanding of participants’ social position according to the ideological assumptions embedded in the questionnaire. Thus, the instruments of knowledge reinforce the disempowerment of participants. The questions reproduce the societal power structures and the social inequalities they seek to investigate. This results in the increased disempowerment of participants who are prevented by their social position and cultural capital from formulating criticism that is politically meaningful. Instead, the participants rely on personal criticisms of encounters with agents of power, such as Housing Commission officers.

New approaches? Some concluding thoughts

We have argued that governing societies requires particular forms of knowledge, and particular ways of knowledge generation support specific forms of governing. The preference for standardised and formalised survey data coincides with the neoliberal model of an autonomous rational actor used to govern societies on the basis of statistical mass data. While this form of governing and knowledge generation has the advantage of transforming the social world into a manageable entity, it is associated with a systematic loss of information and understanding of complexity. Drawing on an exploratory study of the experience of completing survey questions, one element of the creation of statistical data, we have argued that apparently unproblematic survey questions reflect and reinforce dominant social and economic perspectives. As a result, survey respondents are subject to symbolic violence as their experience is denied and discredited (Bourdieu, Reference Bourdieu1999). The long-standing administrative relationships taught residents that their opinions would not be taken into account and were not worthy of consideration. In turn, this experience reinforced their beliefs about political participation, participative urban planning and place-based policies.

At the beginning of this century Hess and Adams (2002: 71) somewhat optimistically observed that in Australian public administration the knowledge frame had shifted from a positivist approach linked to a microeconomic model of policies to a hermeneutic approach linked to the development of social investment and community partnership policies. They suggested that new instruments and skills were required to fill the need for a new type of knowledge. Amin (Reference Amin2005: 620) suggests that the development of new sets of measurement will promote a morally normative frame, thus hindering the development of policies using alternative theoretical approaches. He also considers the moral underpinning of the place-based approach dangerous when it holds communities responsible for their social developments. Boehm's research casts some doubt that a hermeneutic model of administrative knowledge can easily overcome entrenched inequalities and participants’ learnt inability to make a strategic use of knowledge. For residents of an area where long-standing detrimental administrative relationships have shaped inhabitants’ disposition to engage in political participation, new instruments of knowledge may fail to identify individual circumstances that are shaped by societal power structures that hinder political participation. In other words, the research instruments may not recognise or address the processes of misrecognition, and thus will continue to reinforce certain forms of knowledge and ways of seeing (Lister, Reference Lister and Lovell2007; Peillon, Reference Peillon1998). Furthermore, new methodological tools may fail to investigate how knowledge is constructed through long-term social interactions, local administration and the unequal relations of power it reproduces (Bourdieu, Reference Bourdieu1999).

Knowledge generation is not neutral, but it is influenced by social and political relationships that enact and generate knowledge. Thus, new methodological practices for the generation of data will not challenge the framing of social issues or the purposes to which ‘evidence’ are put. In other words, the development of new methodological approaches must recognise the levels of mistrust in government and their social consequences, and address social structures of domination (Bourdieu, Reference Bourdieu1999). While this may cast doubt on participatory knowledge generation, it highlights the importance of qualitative methodologies (for example, action research, ethnography, etc.) that can better recognise the complexity of lived lives and local communities to inform public policies (Hess and Adams, Reference Hess and Adams2002: 69; Young et al., Reference Young, Ashby, Boaz and Grayson2002: 215; Judd and Randolph, Reference Judd and Randolph2008: 99; Denzin, Reference Denzin2009). Nevertheless, we expect that the institutional and organisational practices that rely on measurement and quantification will persist. Survey instruments are based on long-term skills and organisational investment. As the institutional history of statistics shows (Desrosieres, Reference Desrosieres2010), instruments have an independent life and some epistemological practices persist despite their mismatch with the current types of public interventions. The relationship between modes of regulation, the methodologies they use and the knowledge frames they mobilise, and, thus, the social and power relationships they create, invite a pragmatist approach. An open mixed method approach, that considers the long-standing administrative relationships of the participants, could help to reveal what we do not know in an uncertain and changing world.