‘The assessment is in the chat’: Analysing conversations in community care

The research setting

As stated, the data presented here was collected as part of an ethnographic study of a team of social care workers whose job was to co-ordinate community care in a small English town. Nine visits were recorded over the period 1998–2001. The professional participants included representatives of all the occupational groups involved in the team: two qualified social workers, two unqualified social workers, three DSOs (domiciliary services organizers, or home-care organizers), two OTs (occupational therapists) and one nurse manager who participated in a joint visit with an unqualified social worker. All the practitioners were white, all but one, British. Two were men; eight were women. Their ages ranged from 28 to 60 years. The lay participants (‘clients’) comprised nine service users, aged between 52 and 93 years of age, six women and three men (including one married couple, both of whom had care needs arising out of multiple health problems). Six other ‘carers’, relatives of service users, were interviewed by the practitioners, as part of their assessments of need. One visit involved only carers, the son and daughter-in-law of a woman visited previously. All clients were white British.

The research also included participant observation in the team’s work base, follow up interviews of workers and clients in addition to analysing documents. Prior to undertaking this study, the researcher had been a member of the team; during it, he worked as a practice teacher for the authority employing the team and retained his base in the Team Room. None of the visits observed were carried out by students, or former students of the researcher, reducing any direct influence the researcher’s work-role might have had on workers’ behaviour. Nevertheless, the presence of an observer and a recording device inevitably had an impact on the interviews. Arguably, the effect was to make both worker and client more conscious of how other people might evaluate what they said, consequently, workers may have been on their ‘best behaviour’ whilst being recorded. Follow up interviews with clients suggested the presence of an observer may have stimulated some clients to press the justice of their claim for a specific service, whilst making others more anxious not to be perceived as ‘grabbers’, as one of them put it. Recording home visits without the presence of an observer may well have had a similar impact, as it, too, would have exposed participants to external scrutiny. As part of the ethnography, shadowing the visits not only allowed the researcher to get a sense of what usually happened on such visits but also provided valuable points of reference for further discussions. Audio-recording provided an insight into the otherwise invisible work of the home visit; transcribing the recordings allowed them to be analysed in a detail which would not have been possible via written notes, or post hoc reconstructions.

Permission was negotiated with local managers and staff as well as with those responsible for research governance within the authority. Clients were given a written description of the research prior to the visit, via their worker. Oral permission to observe and record the visit was secured before setting out on the visit, again on the doorstep and finally, in follow up visits made without the worker.

The context of assessment visits

A range of external factors affected the home visits differentiating them from the clinically based one-to-one interactions of, say, Silverman’s (1997) counsellors or Ruusuvuori’s (2000) doctors. Among the most salient of these differences were the policy background, the venue of the encounters, which was in clients’ homes rather than in practitioners’ clinics or offices, and the presence of third parties, relatives or professionals, on many of the visits. (Only two of the nine visits comprised a lone professional and their client.) Each of these factors helped shape the pattern of interactions within the visit, but there is insufficient space to discuss them here. Instead, only an outline of the policy background is provided, as this may help compare these encounters with those in other settings.

The visits were carried out after the National Health and Community Care Act (NHSCCA) of 1990 introduced a model of community care in which local authorities were to assess individual need and purchase care packages, tailored to meet those needs, from a variety of suppliers from the public, private and voluntary sectors. There was to be an emphasis on maintaining people at home and reducing costly admissions to residential, or hospital care. For workers in this study, local implementation of the NHSCCA involved a cap on expenditure, which was devolved to team level and regulated by panels of team members. This accentuated their ‘resource husbanding’ tendencies, as predicted by Lipsky’s (1980) model of street-level bureaucracy and reported, elsewhere in Britain, by Ellis et al. (1999). Consequently, much of their practice could be said to be budget-led (Carey, 2003).

For the Team’s clients, the NHSCCA brought uncertainty, especially about who would pay for what (Baldock and Ungerson, 1996), coupled with a threat to existing services, especially housework, which had been traditionally performed by home helps, whose activities were now being ‘re-focused’ to ‘personal care’. Follow up interviews suggested clients may have been unsure about what was at stake and how they should conduct themselves. This means ‘the definition of the situation held by the parties’ (Schegloff, 1993: 114) emerged out of the subsequent interaction, perhaps, to a greater extent than in more familiar institutional situations, such as, doctors’ surgeries.

The NHSCCA can be seen as a preliminary step towards the current ‘personalization agenda’ with its personalized budgets and direct payments. For Team members, it also represented a step towards the computerization of care, but although members had computers on their desks, their assessments were not yet bound by the compulsory screens of computerized assessment protocols (Gorman and Postle, 2003). Instead, workers used paper documents that they modified as they saw fit. As ‘Ursula’ (all names of participants and places have been changed), an OT, said of assessment documents, ‘I’ve just cribbed one together myself, and I’m still changing it every time I see somebody I think of something new to attach to it or ask is something else really relevant anymore because I find the personal profiles are no use at all and not much ornament’. (‘Personal profile’ was the title of an official assessment document.)

Compared to practitioners in other settings (Carey, 2003), workers appeared to exercise more professional discretion about how they carried out their assessments. Furthermore, they did so without explicit guidance on what constituted a need. In Oakeshott’s (1975) terminology, they were ‘teleocratic’ rather than ‘nomocratic’ (Evans, 2012) because the end of keeping people in their own homes had priority over pre-defined eligibility criteria. Instead, workers were expected to maintain a ‘flexible’ approach to need, so long as meeting it helped avoid a costly admission to residential care. As the team manager, Brenda, told her workers, ‘Now the words I don’t want to hear are “we don’t do ironing”, we would say “we don’t think that this is the best use of the home help’s time.”’ On the other hand, discretion was limited by the requirement on workers to justify their assessment decisions to their colleagues, either explicitly in allocation panels or implicitly through an awareness of ‘pressures on colleagues’ (Quentin, social worker). Awareness of such pressures was promoted via the mutual surveillance of collegial relationships within the Team Room.

Agenda setting, turn-taking and topic shifting

Turn-taking, when one person speaks and another takes up the ‘slot’ made available to reply, is the most basic conversational structure (Schegloff, 1993). Often such pairs take the form of questions and answers. In ordinary conversation the participants may take turns to ask questions, however, the ‘chaining’ of questions, so that a question is followed by an answer and then a further question and so on, is a feature of institutional talk, especially interviews (Silverman, 1997). The importance of this simple structure was demonstrated, in my sample, by a contrary example. Workers visited Mrs Rivers following reports that she was not eating or drinking, but Mrs Rivers refused to respond to inquiries about her wishes, needs or intentions other than to repeat the statement, ‘I’ll stop here where I am’. This led to a conversation characterized by the use of a variety of conversational gambits by the workers and long pauses or minimal responses from Mrs Rivers. (Although, at the start of the interview, Mrs Rivers indicated that she agreed to be observed and recorded, the researcher withdrew in case his presence was aggravating her discomfort but the interview was terminated soon afterwards.)

Extract 1: Turn-taking and topic shifting

If workers have power over their clients, this is one of the ways in which it is produced, through the power of the opening turn of an adjacency pair to elicit a response. Mrs Rivers, despite the pauses, eventually replied to some of the questions put to her. The extract also illustrates some of the tactics that may be used to resist such power. The pauses clearly indicate Mrs Rivers’ reluctance to take up the topic she is offered in the opening turn. When she does respond, she deploys a number of devices to shift the topic of the conversation: interrupting to introduce a new topic, shifting a response towards a new topic, declining to take up a new topic and persevering with the previous one. At line 3, Mrs Rivers interrupts Naomi to shift the topic from how she ‘spends her day’ to ‘stopping here’. Naomi, in turn, attempts to shift the topic to ‘pain’, but Mrs Rivers’ response at line 6, may be read as either a reply, followed by a shift back to her previous topic, or as persevering with the topic of ‘stopping here’. Either way, she successfully avoided picking up topics proffered by the practitioners except that of her great-grandchildren. (This topic up-take illustrates a tactic frequently used by workers on home visits the use of ‘props’, such as photographs, to open a line of enquiry). In Mrs Rivers’ case, topic shifting can be seen as resistance, but more generally, it is the process through which the agenda and scope of encounters are negotiated.

One important way of shifting the topic, not used by Mrs Rivers, is to ask questions. Heritage (2010: 1) reports that in medicine, ‘questions are developed in a branching structure through which specific clusters of diagnoses are successively pursued or ruled out in a process of hypotheses testing’. In this study, some workers proceeded in a similar ‘interrogative’ (Smale et al., 1993) manner, as we have observed Naomi and Hazel attempting with Mrs Rivers. Quentin (social worker) also used this method to explore the viability of care plans for Mrs Rivers with her son and daughter-in-law. Other workers (or the same workers at other times) appeared to use either ‘procedural’ (Smale et al., 1993) or conversational styles, either following a set list of questions or responding to the previous speaker, respectively. Both Fiona’s interview with Mrs East and Evonne’s with Mrs Castle followed this conversational pattern at times. As Evonne (DSO) put it ‘I usually try to get them into a conversation […] get people talking, people open up more then and I think you can find more about a person [than] if you are just shopping listing it’. Most workers appeared to agree, using a combination of conversational and interrogative styles and relying on a ‘mental checklist’ rather than a printed interview schedule. Jack’s interview with Mr and Mrs Parks was an exception and may be seen as an example of a procedural style. Whatever their style, workers asked more questions than clients. Across all nine interviews, two-thirds of all the questions recorded were asked by workers.

The following example, illustrates some features of questioning by workers on home visits.

Extract 2: Types of questions

First, Fiona’s questions in both lines 5 and 9, act as topic shifts moving the interview along Fiona’s agenda. Secondly, Fiona appears to be following a procedural style using a checklist she has internalized. Accordingly, as described by Silverman (1997), questions were frequently asked as a sequence or ‘chain’ of apparently unrelated topics. This questioning style seemed to be common in the early stages of an assessment. Thirdly, questions were frequently asked in pairs as in line 9, where an open, scoping question is followed by a more specifying ‘polar’ question (Heritage, and Raymond, 2012) demanding a ‘yes’/‘no’ answer. This questioning strategy sequentially opened and then focused ‘explanation slots’. Finally, questions could be asked positively, so that ‘yes’ is the conversationally preferred answer, neutrally, as in line 5 above, or negatively, inviting the answer ‘no’. Often, workers’ questions were framed in terms of ‘managing’ or ‘struggling’: ‘How are you managing to get washed and dressed? Are you struggling?’ (Evonne, DSO). Questions phrased ‘do you struggle with?’ or ‘how do you manage?’ can be heard as inviting a statement of need, whereas those phrased ‘you manage okay?’ (Theresa, DSO) can be heard as suggesting there is no need to be met. Overall, workers’ questions appeared to be productive in eliciting information relevant to the purposes of the assessment. On occasion, clients could be heard making ‘statements of need’, i.e. when they described a difficulty in their lives or requested a service such as help filling out a form. In fact, over half (115 out of 224) of such statements were produced as responses to workers’ questions.

In contrast, users and their carers deployed questions in a wider range of uses including organizing the ‘domestics’ e.g. ‘Would you like a cup of tea?’, or ‘tag’ questions seeking support such as ‘Didn’t he Don?’ Only 60 out of 332 questions asked by clients could be coded as directly seeking information relevant to the central task of the visit; further, two-thirds of these were asked by relatives rather than service-users. This suggests that service-users may have felt reluctant to engage with ‘consumerist’ topics of price, availability and suitability and making them more like ‘passive’ recipients rather than ‘active’ consumers of care (Baldock and Ungerson, 1996).

There were, however, passages where clients can be seen to be asking ‘consumerist’ questions. Examples include: Mrs Brown’s ‘How much are stair-lifts?’ and ‘How long does it take for this?’ and Mrs Castle, who asked, with the conversational assistance of her daughter-in-law, ‘If I could have somebody Saturday and Sunday?’ A reluctance to ask questions that could be heard as unambiguous requests for services extended to questions about entitlement, ‘welfarism’ in Baldock and Ungerson’s (1996) terms. An unambiguous example of such a welfarist question came from Mrs Castle’s daughter-in-law, ‘Is she allowed to have any more?’ Again, such questions were more likely to be asked by relatives than by service-users themselves, and contra to Baldock and Ungerson’s (1996) individualist/collectivist axis, those who asked consumerist questions seemed also likely to make welfarist enquiries about entitlements.

Workers asked more questions than their clients and therefore, did more to set the agenda of the interview. Thus, to some extent, they could be said to have greater control over it. Clients, in turn, influenced the course of the conversation by their responses to questions. As we have seen, refusing to take up topics proposed in a question could be a means of resistance. Clients could also change the topic by extending their answers to express their own concerns. For example, Mrs Brown answered Ursula’s question, ‘Can you manage getting on and off the bed?’ With, ‘Yeah, I’ve not got arthritis’. Later, she told Ursula, that her arthritis was cured through ‘Divine Healing’ an event that seemed significant to her sense of who she was and how she managed to live with an uncooperative body.

Frequently, clients’ topic shifts seemed to extend an answer to address their moral adequacy as may be seen in Extracts 3 and 4.

Extracts 3 and 4: Topic shifts and moral adequacy

Both examples follow Wooffitt’s (2005: 105) ‘I was doing X, when’ format. Wooffitt argued that this construction helps establish the teller as an ordinary person to whom transforming events such as widowhood or illness have happened. The implication being that without such external events both Mr Parks and Mrs Dove would have remained ‘independently minded’ as Mrs Dove put it.

One of the most prominent means by which clients shifted a topic raised by a worker’s question was by telling stories. Extract 3 forms the start of a long narrative told by Mr Parks of his medical history, whereas Extract 4 can be read not only as a short biography in itself but also provides the start of an account of why Mrs Dove needed a stair-lift. After describing a series of operations she concluded: ‘me home help, she saw the state I was in, “right”, she said “you need a stair-lift”. It were her what put in for it not me’. In the conclusion of her story, one can hear Mrs Dove distancing herself from asking for a service and compromising her moral adequacy. Most clients told stories in responses to questions or proposals (see below) from workers. Their tales could be classed in a number of genres: general biographies, medical stories, reports of domestic routines, explanations of how they came to receive a particular service or accounts of their experiences of services. Storytelling allowed clients to take control over sections of the interview, temporarily at least, setting the agenda to account for their moral adequacy and to express their needs.

Approximately, a third of all statements of need were produced via stories, such needs might be ‘consequential’ (Bury, 1991) i.e. have practical consequences for the person’s care, as in Mr Mason’s story about asking a nurse if he could have a bath. Other stories described what was important to that person or, like Mrs. Dove’s story, what sort of person they were. This is ‘significant’ meaning in Bury's terminology. The formulation ‘I was X, so’ forged a link between the kind of person the subject of the story is shown to be and their need for a particular type of service. For example, when Fiona suggests to Mrs East that they visit a day centre together, she responds by telling a story about her childhood deafness, which led to her becoming ‘a bit of a loner’ and, by implication, not in need of day care.

Introducing, or shifting a topic provides a measure of the power to set agendas, but this is only one side of the equation, how a topic is received is also significant. Responses to stories could pick up on significant meanings, as in Jack’s ‘So that changed a lot of things’, or Fiona’s ‘And they do miss him’, but they rarely made an explicit link between the need told in the story and a specific service. The content of stories, however, could be used by workers to guide their inquiries and construct a care plan. Quentin did this when he drew upon Mrs Rivers’ character as revealed by her son’s stories, to outline the difficulties of supporting her with home helps. On occasion, a worker appeared to dismiss a story of how a client was experiencing difficulties by summing up, ‘But you were managing’. Often, workers seemed to ‘heckle’ (Ruusuvuori, 2000) using interruptions to steer the conversation back to their agenda. For example, when Mrs Brown told a story Ursula appeared to wait for a break in the flow to pop in her next question. Sometimes she picked up on a minor component of the story to move onto the next question, at others she seems to be referring to a mental checklist.

As we have seen, Lipsky (1980) claimed that workers interrupt their clients in order to control them. As with questions, however, interruptions have to be understood by reference to the work they do in their particular conversational context (Schegloff, 1993; 2007; Wooffitt, 2005). They took different forms and seemed to achieve different purposes depending on when they were made and who made them. Some interruptions acted as topic shifts while others, marked by sequences of over-lapping speech, suggested that the topic was of interest to the participants (Tannen, 2001) rather than contested attempts to dominate the conversation. Contrary to the findings cited by Lipsky, users and carers interrupted workers more frequently than vice versa. Further, they used interruptions in different ways. Workers tended to interrupt to chivvy the interview towards a conclusion. In contrast, interruptions by users and carers often seemed to indicate reservations about the proposed course of the assessment. For example, Mrs East interrupted Fiona’s talk of benefits to raise the topic of adaptations to her home:

F: So we know what to put really//

As discussed above, such interruptions could shift the topic by telling a story explaining why something was necessary, for example, Mrs Rivers’ son apologized for interrupting Quentin to tell a story of his mother’s previous experience with home helps. Overall, thirty-nine statements of need were produced in this way. Richards (2000) also noted the use of interruptions by people undergoing community care assessments, arguing they signified a form of tentative resistance by clients to plans proposed by workers.

Discussion: Conversation and models of social care

One of the aims of this article was to explore whether the use of topic shifts could be used to explore interactions between workers and clients in social care. Clearly, Schegloff (1993) is correct to contend that features such as interruptions have different effects according to their place within a conversation. This makes claims such as Lipsky’s (1980) assertion that street-level bureaucrats use interruptions as a means of controlling clients open to question. Nevertheless, it was possible to distinguish questions, or interruptions, which set, or shifted, the topic of conversation from other uses of these devices and to observe how they were used to set the agenda of an assessment. Not surprisingly perhaps, workers’ greater use of questions suggests they were responsible for guiding the content and pace of interviews. For their part, clients were more likely to introduce topics through extending their answers to workers’ questions, telling stories and interrupting workers when plans were proposed.

Returning to Smale’s (1993) models of assessment, workers could be observed asking questions diagnostically, as in the questioning model, and also from a schedule, as in the procedural model. None of the interviews, however, fitted the exchange model with a symmetrical use of questions by both parties. Given clients’ uncertainties about the scope of the assessment and the possible costs of care, it may be that they allowed workers to define the parameters of the interview. Silverman (1997) reports a similar process for counsellors and their clients. There were, however, passages where service users, or their relatives, could be seen as asking consumer orientated questions about a plan proposed by the worker. Nevertheless, storytelling seemed to be the principal means by which service users made their needs and preferences known. Narratives also provided a medium for establishing clients’ moral adequacy. For service users, this appeared to be related to the delicacy with which asking for help was associated. Taken together, these factors suggest the exchange model would require attentive nurturing if it were to become a reality.

Workers in this study fitted, at least to some extent, Lipsky’s (1980) model of street-level bureaucracy. Backstage observations made it clear that they not only had an interest in husbanding resources but also judged the ‘genuineness’ of clients. In particular, workers were critical of people who appeared to them to have elected to become dependent (Ellis et al., 1999). Such concerns were, however, less obvious in workers’ face to face interactions with clients. Here, workers sometimes denied being influenced by resource considerations and appeared to avoid giving any indication that they disapproved of a client’s actions. Nevertheless, the variations in phrasing questions and responding to stories, outlined above, suggest workers were able to either encourage or discourage clients to formulate specific needs. It is here that workers’ discretion may lie. It might be argued that it is the outcome that counts and that the way questions are phrased is simply a matter of style (Howe, 1991), but out of such details cases may be thickened, or thinned (Pithouse, 1998), according to the worker’s preoccupations. As Evonne put it: ‘You start your assessment from the moment you pull up outside, you know she likes it kept well’. Workers recognized this scope for discretion, one social worker, commented that some of her colleagues would do ‘anything for anyone.’ Others would not, ‘Unless it was one of their favourites’. Then they would go to ‘the ends of the earth. Otherwise, it was straight down the middle “it’s my job”’.

Recording the interactions between workers and clients allows us to look more directly at practice than would otherwise be possible. Even in a small sample, the picture revealed shows considerable variations in the form and content of the interactions. Partly, this may be an effect of looking closely at practice, but it also reminds us of the role clients played in shaping assessments, a process that may have been accentuated by clients having ‘home-advantage’ so to speak. Lipsky focused on workers’ attempts to control the behaviour of their clients. In this study, we have seen that such attempts are not always successful and that even if workers set the overall agenda, in line with official priorities, their attempts may be resisted or subverted by their clients. Rather than thinking in terms of the interactions fitting a particular model, or workers following a particular style, it may be more useful to think of both workers and clients using a repertoire of conversational moves and responses to negotiate what counts as needs. And if street-level bureaucracy seems closer to actual practice than the exchange model, it may be useful to consider the equality and mutual respect, proposed in that model, as an aspiration rather than a description. If the scope for the exercise of discretion lies in the conversational details of practice, and practitioners aspire towards promoting such active citizenship, they may well profit from a closer study of how everyday practice with clients is accomplished.