Street-level studies of activation

Dunn is right to call for opening up the employment policy debate to the views of other actors, particularly those at the front-line of activation policies, but he is not correct when he says: ‘it is noteworthy that social policy authors have drawn conclusions about unemployed claimants without considering the views of agency workers they come into contact with’. While it is the case that much evaluation of active labour market policies is driven by quantitative methods and econometrics, there is a rich literature on policy implementation that places the agency of employment service workers front and centre in both theoretical and empirical studies. There are examples from the UK and elsewhere in Europe, as well as the US and Australia that analyse the views of workers in drawing conclusions about unemployed claimants.

Sharon Wright's (Reference Wright and Millar2003) work on employment services staff in the UK, for example, paid close attention to the attitudes of staff towards the unemployed clients they worked with. Wright's research showed that staff had a variety of ways of categorising clients, some of which conformed to dominant cultural attitudes towards the unemployed. Elsewhere in Europe, in countries such as the Netherlands, Sweden, France and Denmark, there are some exemplary studies of worker attitudes and worker's agency in implementing active labour market policies in an era of neo-liberalism and managerialism, particularly in regard to the performance management targets that constrain the work of employment advisers (Thoren, 2008; Dubois, Reference Dubois2010; Van-Berkel, Reference Van-Berkel, Betzelt and Bothfeld2011). Much of this work is informed by the street-level bureaucracy perspective developed in public administration and political science, pioneered in the classic study of Michael Lispky (1980).

More recently, US researchers into active labour market studies have revived and further refined this conceptual lens. For example, there are sophisticated qualitative studies of how African-American welfare bureaucrats both challenge and sometimes adopt racist stereotypes in their work with unemployed families (Watkins-Hayes, Reference Watkin-Hayes2009), there are ethnographies of the employment office and how workers actively deter people from making claims on the state (de-Parle, Reference De Parle2004) and mixed methods studies of how workers apply sanction in ways that reinforce race and class divisions in society (Soss et al., 2012). These are just some examples of work that include an analytical focus on the attitudes and actions of front-line activation policy workers. Some greater acknowledgment of this body of work would have been welcome in Dunn's article.

The main finding of the empirical section of Dunn's piece is that workers who work with unemployed people subscribe to the view that a ‘dependency culture’ exists among certain groups of unemployed. The study respondents were also reported as believing that the unemployed are ‘choosy’ in the jobs they are willing to take (which is close to saying that unemployed people choose their unemployment). These observations are based on a study where Dunn interviewed forty people employed by agencies who work with unemployed people. It is not remarkable or surprising that these workers believe a ‘dependency culture’ is a characteristic of being unemployed. The ‘dependency culture’ thesis is a dominant discourse in the more liberal welfare states where successive governments have adopted a punitive and individualistic approach to managing poverty and its consequences. Dunn's analysis here would benefit from an historical account of the origins of ‘welfare dependency’, as this would help explain its hegemonic status and hence its appropriation by workers (Fraser, Reference Fraser1997).

The assessment tools, the case management approach to employment services and the sanctions that these workers are expected to use reflect a moral framework that aims to discipline subjects into a pre-configured conception of ‘activation’. In light of this, we should probably be more surprised when we find examples of front-line workers who refuse to adopt this dominant position. Indeed, there are published studies where front-line workers refuse to cooperate, or where they struggle with the ethical dilemmas of enacting a professional ethic of social justice or care alongside what they consider to be a policy that is unjust (McDonald and Chenoweth, Reference McDonald and Chenoweth2007). As Dunn notes in his article, it is this kind of oppositional ethical standpoint that many social policy researchers in the UK and elsewhere have also adopted in their critique of activation policies

The duality of agency and structure

Dunn is concerned that in dismissing the culture of poverty thesis of the conservative welfare reform theorists, such as Charles Murray and Lawrence Mead, the UK social policy academy has avoided any notion of individual agency in explaining unemployment. In my reading of this critical social policy literature, the objection is not so much to the existence of sub-cultures or the possibility of agency, but that sub-cultures or bad personal choices are routinely put forward as the principle causes of unemployment. The concern is that the social problem of unemployment has been turned into the individualised problem of the unemployed.

Moreover, most social policy scholars do not seek to ignore individual attitudes and motivations, but these are seen as one of the many inter-connected issues that shape people's experience of unemployment and pathways to employment (Berthoud, Reference Berthoud2011; Shipler, 2009). There is also a concern among many in the mainstream social policy research community that cause and effect have become confused and conflated in dominant policy frames about unemployment. Feeling despondent in the face of rejection by employers, anger in the face of bureaucratic indifference or feeling a sense of shame about living in poverty can be understood as predictable lived effects of long-term unemployment, particularly in a society that eulogises consumption and the paid-work ethic as principle markers of citizenship. Indeed, there is a great deal of scholarship that highlights the harmful effects of unemployment on social and psychological wellbeing (Sennett, Reference Sennett1998; Carroll, Reference Carroll2007).

Moreover, it is possible to accept the notion of personal responsibility while still insisting on structural explanations for injustice and inequality. In other words, the response to the social problem of unemployment does not need to be an ‘either/or’ proposition in terms of agency and structure, clearly it can be an ‘and/also’ concept as urban sociology illustrates (Bourdieu and Wacquant, Reference Bourdieu and Wacquant1992). Mono-causal explanations of complex problems like unemployment are simply implausible. Indeed this is one of the major problems with Charles Murray's thesis on poverty and unemployment. He sets up a false dichotomy between structural causation and personal responsibility. Implicitly, Murray also assumes that policy makers and citizens need only worry about the responsibility of the poor; members of other social segments, for the most part, act responsibly (Marion Young, 2011). To a certain extent, Dunn's review of Murray's work and of UK social policy scholarship on unemployment reproduces this dichotomy between structural causes of unemployment and personal responsibility.

Avoiding this dichotomy means acknowledging the background conditions that many unemployed people face, while also acknowledging that individuals will choose different responses to these circumstances. To deny these background conditions, such as poor schools, dilapidated housing, few local employment opportunities is to assume that these background conditions are not unjust (Marion Young, 2011). Is it really possible to claim that societal institutions, businesses and government are doing all they can to remove barriers for everyone in society to achieve economic wellbeing?

There is a tendency in the literature review and the empirical discussion to cite academic literature that reinforces an individualistic explanation. An example of this is when Dunn is discussing the worker's views that older unemployed people are considered less likely to ‘want employment’. Dunn cites secondary research that suggests older people have ‘done their stint’ and are therefore absolved from negative moral judgement. While this may be the case, there is also considerable evidence that older workers face significant age-based discrimination in labour markets in countries such as the UK, the US and Australia (Standing, Reference Standing2011; Gregory, Reference Gregory2001). This experience of discrimination, which can lead to discouragement may also explain why older people were perceived by these workers in Dunn's study as a group that ‘didn't want employment’.

A dependency culture

Dunn moves on from general observations to discuss the consensus view among the respondents in the study. This part of the discussion is focused on a ‘dependency culture’ existing among clients. This belief is reported as based on a range of observations by the workers, including similar negative attitudes to employment within the same family and considerable knowledge of the ‘welfare system’. The other major area of respondent consensus discussed in the article was that ‘significant numbers of long-term JSA claimants they worked with were being too “choosy” in the jobs they were willing to apply for’. In reporting this finding, Dunn states that: ‘Clients were said to be unwilling to apply for jobs they considered inadequately paid, uninteresting, unpleasant, inconvenient (mainly in terms of distance from home) or unsuitable.’ It is at this point that I wanted to know some more about the circumstances of the workers themselves, particularly their employment conditions, such as case load, performance targets and what resources they had available for training and job placement – particularly under ‘quasi-market’ models of employment service delivery. Are the clients considered too choosy from the perspective of the workers who themselves may be under considerable managerial pressure to place clients into jobs, of whatever quality and of whatever duration? These background organisational conditions do not feature in the discussion and as a result the explanation of worker attitudes remains incomplete. Another part of the context that is missing is a spatial analysis, particularly how geographical differences have a significant influence on the implementation of activation policies (Soss et al., 2012; Whitworth, Reference Whitworth2012).

I was also left wondering whether the strong endorsement of the ‘dependency culture’ also reflected a ‘researcher effect’ in the same way that Dunn believes operates in studies with unemployed people who indicate a strong identification with the work ethic. Shouldn't this same qualification apply when asking workers whether they think a ‘dependency culture’ exists among groups of unemployed people? Dunn acknowledges that the advisers in his study overestimate the extent of intergenerational unemployment when compared with government statistics on intergenerational unemployment. As Dunn acknowledges, the workers may overestimate because they are working with more disadvantaged unemployed clients and so this comes to influence their view of all unemployed clients. Perhaps they were also influenced by the choice of questions or more subtly by the tone of voice of the researcher, all of which can influence the subject into giving answers skewed towards the interviewer's own opinions, prejudices and values (Al Natour, Reference Al-Natour2011). Like all researchers, Dunn needs to be mindful that he is not using leading questions to direct responses. One can imagine getting quite a different response to a question that is framed with the assumption that some ‘clients did not want to work at all’ compared with an interview that starts by questioning whether this assumption is warranted.

The language and key concepts used in interview questions are also likely to influence responses. There is a big difference between the concept of being choosy and the notion of having a choice. This distinction matters not just in terms of ensuring concept validity but also in terms of the implications of the findings. ‘Choosy’ is a loaded term when interpreted as a negative character trait of someone who is being too picky, whereas ‘choice’ has positive connotations within a liberal philosophy of rights and freedoms. Moreover, being ‘choosy’ goes against the contemporary cultural grain, as it indicates that someone is unwilling to move to take up a job or reinvent themselves in an age of ‘flexibility’ (Sennett, Reference Sennett2006) where the ideal worker citizen is someone that is entrepreneurial with few meaningful attachments to place or people. To be ‘choosy’ then is not only being irresponsible, it also means the person may be seen as rigid in their wish to pursue only certain types of employment or for an individual to decide that the costs simply outweigh the benefits. This choice may be perceived as irresponsible, but it may be entirely rational and reasonable to the person who must live with whatever choice is made.

In discussing the implications of these worker attitudes, Dunn is right to acknowledge that it is hard to separate one's political beliefs and moral principles from what we research and how we report what we find. Some social policy scholars will still find themselves opposing a policy on the basis that they find the means beyond the pale (e.g., sanctions, prescriptive activation requirements) regardless of whether a particular end is achieved (employment). Other scholars might object to the narrow and economistic conception of welfare and wellbeing embodied in these policies, such as the de-valuing of care in the formation of social citizenship (Williams, 1999). Other scholars might accept that some forms of state paternalism are justifiable where there is a clear public interest being served, but they might also conclude that in the case of welfare-to-work type policies passing the public interest test is far from clear.

In drawing his conclusions about these matters, Dunn seems somewhat ambivalent. He emphasises the importance of skill development, but he is also sympathetic to the idea of ‘sticks’ when he states on the basis of the findings that ‘the consistency of the views expressed by respondents must surely go some way to dispelling the possible arguments that JSA claimants’ employment commitment is so strong that attaching job search conditions to the receipt of their benefits is unnecessary’. To be confident of this conclusion would depend on making a well-rounded assessment of all the available evidence, not only the findings of interviews with forty workers. Moreover, simply because a sample of workers believe a culture of dependency is widespread, it does not follow that this justifies current policy parameters – as to do so would conflate the views of workers with the views of the unemployed. Indeed, some workers probably need to be challenged on their views about the unemployed, in the same way that employers need to be encouraged to employ older workers, people with disabilities or single parents. Arguably, the disciplinary welfare gaze needs to be turned on its head and focused back towards the state, employers and front-line workers. And in explaining why these workers might have these views we should look at the conditions of their work, their caseloads, their employment conditions and the level of resources they have at their disposal to invest in people to help them make the transition from unemployment to a decent job.

The effectiveness of sanctions

Overall, we will need to remain cautious about assuming that sanctions and conditions are an effective method of facilitating behavioural change (regardless of any moral objection). It is certainly the case that authors like Mead and Murray believe that people need ‘help and hassle’ in order to move them into employment. However, this conviction alone does not make for good employment policy. Arguably, a system that is based on discipline and control, that requires activity for its own sake, which diminishes the scope of people's autonomy will not support the development of long-term employability (Fowkes, Reference Fowkes2011). Moreover, there is little evidence that sanctions change work-related attitudes (Mitchell and Woodfield, Reference Mitchell and Woodfield2008). Recent evidence from Australia suggests that people subject to these policy measures have an overriding sense of the ‘hassle’, but little sense of the ‘help’ (Murphy et al., Reference Murphy, Murray, Martin, Chalmers and Marston2011). Alford and O'Flynn (Reference Alford and O'Flynn2012) make a different but related point about the impact of sanctions on self-efficacy and motivation:

Perhaps at this point in the article, it would have been useful for Dunn to briefly reflect on the growing international evidence about the effectiveness of ‘work-first’ and workfare type policies. In evaluations of the UK's ‘work-first’ activation regime, it has been found that the programme represents poor value for money and there are substantial deadweight effects (NAO, 2010). In Australia, there is independent evidence that ‘work-first’ policies have only benefited a small percentage of the unemployed and most of those affected by policies are economically worse off (Davidson, Reference Davidson2011). Martin and Grubb (Reference Martin and Grubb2001) came to a similar conclusion in their earlier study of active labour market programme impacts across OECD countries where they found that the results were not very encouraging in terms of job placement, particularly for young unemployed people. They did find that there were some positive results in that wage subsidies and labour market training did work for some groups. However, other studies have found that wage subsidies and similar measures can have a displacement effect (Calmfors et al., 2002). Other researchers are critical of the level of analysis that much active labour market research operates at, arguing that micro-economic and macro-economic studies tend to treat programmes and interventions as simple, ignoring the often complex contexts in which these programmes operate (Funnell and Rogers, Reference Funnell and Rogers2011). What is missing from the conclusion to Dunn's article is what the study means in terms of the obligations of the state and the agency of government and business actors. Surely, if the state makes an argument for strong paternalism, in that it dictates both the means and ends of employment policy, then the onus is on the state to show that its ‘tough love’ approach is working.