The use of Facebook in social work practice with children and families: exploring complexity in an emerging practice

Abstract This article draws from a 15-month participant observation study of social work and child protection practices in England to illustrate how social workers used Facebook to gain another view of service-users’ lives. Social media use was not an intended focus for the study, its presence emerged during our data analysis. While some research has shown that such practices occur, our long-term ethnographic approach provides new insights into how Facebook was actually used in ongoing casework with families and why it was used. Our findings show that Facebook use took multiple forms. Some social workers actively searched service users’ Facebook pages and some opposed any such usage. We further advance the literature by introducing a third group who were unwillingly “drawn into” acting on Facebook information presented to them by others such as their managers. Our research insights suggest that social work must pause to consider the implications of these complex emerging practices.


Introduction
Social media has redefined how we are able to keep in touch with, locate and relate to individuals and communities globally. Social networking sites have provided increasingly simple methods for people to share and access rich information about each other's lives using mobile devices untethered by restrictions of time and place. This social media practice of publishing aspects of daily life to a wide community is often referred to as a "collapsing of borders" (Marwick, 2012;Marwick & Boyd, 2011;Trottier, 2011). Kolmes and Taube (2014) compare this collapse of borders to a "small world effect" whereby everyone knows everyone and their business.
There has been a growth in literature concerned that social workers are viewing service users' Facebook accounts to gain another view of their lives, apparently without their knowledge (Ballantyne, Lowe, & Beddoe, 2017;Stanfield & Beddoe, 2016; British Association of Social Workers (BASW), 2018). This article draws from a 15-month participant observation study of social work and child protection practices in England that aimed to find out how social workers established, developed and sustained relationships with service users in long-term casework. The use of social media was not an intended focus for the study but its presence emerged during the fieldwork. While existing research has used surveys, interviews and focus groups to reveal that social media platforms are used as part of child protection work, our long-term ethnographic approach provides new insights into how Facebook was used in ongoing casework with families, and why it was used. This research is important because as Sage and Sage (2016a) observe with reference to service user assessments, there is a lack of research on how social media is being used to inform practice decisions. Trottier (2011), discussing surveillance studies in general, notes that: "social media make large sections of social life visible, and investigative agencies are taking advantage of that visibility. Surveillance studies need to focus on how this visibility is being used by these agencies" (p. 6). We aim to build on existing research by presenting longitudinal, observational as well as interview data about the use of Facebook in long-term social work practice with children and families.
At present, Facebook is the most popular social media platform in the world, with over two billion monthly active users (Statista, 2017). Our findings suggest that it is being used by some social workers and managers who may choose to view service users' Facebook accounts, while some others may choose not to do so. Our analysis reveals another Facebook practice that occurs due to the actions of third parties, where social workers and managers are "drawn in", thus removing their choice in whether they engage or not with this platform. This paper raises a number of challenging issues related to the use of Facebook in social work and child protection. While not the first to raise awareness that Facebook is used as part of child protection work, our longitudinal and observational data brings its use to life in new ways and suggests that much greater clarity is needed on the legal and ethical use of social media information by social workers. Rather than simply criticizing social workers' questionable use of social media, our aim is to try to understand it by making an original contribution to debates about the complex ethical (and in England, legal) issues social workers and agencies face when using Facebook information as part of managing risk in child protection work.

Brief note on terminology
After our findings emerged from the data, we initially analyzed social workers' decisions about social media using the lens of two ethical approaches, Kantian and Utilitarian (Gray & Webb, 2010). We became aware that what we were writing about was much more complex than this binary analysis suggested especially when some social workers felt they were "drawn in" to having to use Facebook. Thus, when we use the term "monitoring" we are referring to instances where social workers make service users aware that they are observing their social media accounts. In contrast to this position, we use the term "surveillance" where social workers may, for example, access a parent's social media account for the purpose of covert surveillance. Being "drawn in" refers to situations where practitioners face the dilemma of whether to look or not at a service user's online presence based on Facebook information sent to them by third parties, including their managers. This condition of being "drawn in" has not previously been identified in any of searches of the literature.

Literature review
To place our research in context, a literature review of social worker use of social media in practice was undertaken. We searched for international research on social work and social media published in English and found 100 articles which variously addressed social workers' use of social media, and the ethical aspects of use in practice. Our search revealed a relative paucity of research specifically reporting social worker access of service user pages (n ¼ 8 articles). To extend our understanding we also examined research conducted with social work adjacent professions, such as counseling, psychology and psychiatry (n ¼ 19 articles). Three main topics were identified from the literature review: the specific use of social media in child protection social work; variability in how codes of ethics and policies are dealing with social workers' and/or their agencies' use of social media; and the "pros" and "cons" of service user monitoring and surveillance within the helping professions in general. Each of these areas is now explored to provide a background to our findings and discussions.

Social media and child protection social work
There appear to be distinct differences between how social workers in child protection think about and use service user surveillance in contrast to social workers in other fields, particularly those in mental health (Ryan & Garrett, 2017). Based on practitioner interviews, Ryan and Garrett (2017) note that child protection social workers are "more open [than others] to using Facebook to gain client information" (p. 6), concluding that "this highlights lack of consistency in social work and, significantly, despite sharing the title "social worker" and being apparently committed to the same ethical codes, stark differences existed" (p. 10). In a more extensive survey of social workers in child protection, Sage and Sage (2016a) found significant levels of comfort around service user viewing. For example: 59% indicated that they had searched for a missing client on Facebook; 54% searched Facebook to explore risk factors; 55% believed that it was "acceptable in some instances" to conduct a curiosity search on Facebook for clients, with 43% indicating that they have done so and a further 13% indicating that they did this either daily or multiple times in the day. Only 20% stated that they had never used social media to look for client information. It was thought acceptable in some instances by 65% of participants to search Facebook for a client as part of an assessment. Almost half thought that it was acceptable in some circumstances to create a fake Facebook account, containing no personal information, to interact with clients (see also Breyette & Hill, 2015). In considering the conditions that lead to child protection social workers conducting service user surveillance, both Stott, MacEachron, and Gustavsson (2017) and Sage, Wells, Sage, and Devlin (2017) have recently examined the role that managers respectively play in this practice. Stott et al. (2017) for example looked at US-based training of social workers and found that 64% of the agencies surveyed offered administrative training on how to both search for and communicate with clients on social media. However, despite this, only 21% of agencies indicated that they had a specific policy governing social media use. Stott et al. (2017) concluded "that respondents believed, overall, that the benefits of social media use by staff far outweigh the risks" (p. 229), indicating that, at least in child protection in the US, there is a high level of comfort around, and even expectation of, the use of social media by social workers in child protection. In another US study Sage et al. (2017) reported on the attitudes of supervisors and that 51% of social workers who had supervisors who approved of social media surveillance searched for a client, compared to 32% of those without supervisor approval. Clearly then, the attitude of the supervisor and culture of the organization have an influence on the subsequent Facebook monitoring or surveillance behavior of any individual social worker.

Codes of conduct, ethics and social media
Social media "is a contested space, highlighting many challenges, most of them falling squarely within the realm of ethics" (McAuliffe & Nipperess, 2017, p. 131). A common criticism is that social work is unable to keep up with the rapid changes in social media and their ethical implications. It is noted that many codes locate service user surveillance in a "grey" area with most claiming that each individual situation has to be considered on its own merits (BASW, 2018; Canadian Association of Social Workers, 2014).
For example, in England where our research took place the BASW Social Media Policy (2018) highlights the need for practitioners to respect appropriate personal and professional boundaries. Because this policy is very closely linked to BASW's Code of Ethics (2014), practitioners are expected to use these to work out the professional practice implications of social media use based on the individual circumstances presented. The Association of Social Work Boards (2015) in the US recommends the use of informed consent for service user monitoring. The Australian Social Work Association (2016) offers extensive guidance and Reamer (2017) suggests that social workers must only conduct non-consented searches of this nature in emergency situations. In contrast the UK Health and Care Professions Council (HCPC, 2017) and New Zealand professional bodies make no explicit mention at all of service user surveillance (Aotearoa New Zealand Association of Social Workers (ANZASW), 2013; Social Workers Registration Board (SWRB), 2014).
Where there was an absence of codes and/or policies, researchers suggested that these be developed (Boddy & Dominelli, 2017;DiLillo & Gale, 2011;Doel et al., 2010;Mishna, Bogo, Root, Sawyer, & Khoury-Kassabri, 2012;Reamer, 2013aReamer, , 2013bReamer, , 2017Ryan & Garrett, 2017;Stott et al., 2017). However, two articles argue that even when they exist, ethical codes and policies alone are never enough to curb ethically questionable practices (Sage & Sage, 2016b;Voshel & Wesala, 2015), and that individual practitioners still need to exercise personal judgement. Sage and Sage (2016b) emphasize that "policy cannot replace critical thinking, clinical considerations, or address all ambiguous practice situations" (p. 83). This point is underscored by Ryan and Garrett (2017) finding that none of the government-employed social workers they interviewed were aware of the government guidelines for social media use, and some asserted that service user surveillance is not acceptable practice. However, even when social workers take this position and believe that this practice is not acceptable, they may decide that the benefits outweigh the costs, as articulated by a social worker: "I have searched for clients on social media, especially when a client runs. I feel this is unethical, but will continue to do so, and I believe many child welfare workers feel the same" (Sage & Sage, 2016a, p. 104). This pragmatism is suggested by Baker, Warburton, Hodgkin, and Pascal (2014, p. 470) when they ask their readers to consider "for example, the likelihood of effective assessment … if a social worker does not include aspects of a person's networked life".

Ethical perspectives
Kolmes and Taube (2014) note, using an analogy of a practitioner physically following a service user, compared to following them around online, that "the crucial differences between this analogy and the intentional searching for client information on the Internet are the ease, convenience, invisibility, and inexpensive nature of the activity" (p. 8). This ease of access to information creates "pros and cons" positions on usage.
The literature reports the "pro" position, which addresses service user safety, emphasizing the ends justify the means (Clary, 2014;Clinton, Silverman, & Brendel, 2010;Kolmes & Taube, 2014;Mearns, Richardson, & Robson, 2015;Sage & Sage, 2016a;Voshel & Wesala, 2015;Zur, 2012). For example, a practitioner might choose to look up the profiles of parents, siblings and/or caregivers, and even potential foster parents associated with a young client to assess the client's present and ongoing safety (Jent et al., 2011;Ryan & Garrett, 2017). Echoing Baker et al. (2014), Sage and Sage (2016a) note, in "a risk-focused orientation, child welfare workers should be thorough in their family assessments, exploring any resource available, including social media. Client privacy and confidentiality is seen as secondary to child safety from this lens" (p. 96). Social media searches are, in effect, contributing to assessments (DiLillo & Gale, 2011;Poloncarz, 2014;Ryan & Garrett, 2017;Sage et al., 2017;Sage & Sage, 2016a).
On the "cons" side of the ledger in the literature, the focus is less on covert surveillance, but rather the potential risk to a collaborative or therapeutic relationship with service users. Lehavot, Barnett and Powers (2012) stress that the risk of damage to a therapeutic relationship is very high as, even with consent, the practitioner must be careful, due to how (in)accurate the information might be, and how subsequent interventions might pose serious risks to the relationship. Social media has a performative aspect which can lead to posts which exaggerate or misrepresent the truth (Boyd, 2007(Boyd, , 2011 questioning the "reliability" of online findings (Boyd, 2007;Ryan & Garrett, 2017) thus casting doubt on the veracity of this information in informing assessments or with regard to decisions about service user care.
Alongside these concerns, the framing of the use of Facebook as stalking contributes to uneasiness about service user surveillance. Sage and Sage (2016a) found that 43% of the social workers they surveyed had searched for service users out of curiosity, with half saying that colleagues had also done so. Further, 55% stated they thought that, in some situations, it was acceptable to conduct a curiosity search for service users. This statistical finding is backed up by other interview data where social workers talking about the aforementioned "creeping" were reporting on how they use social media within their practice to "see if there was congruency between the online client and the face-to-face client" (Clary, 2014, p. 73).
How far client searches stretch confidentiality beyond looking for very clear risks to safety, was a concern (Clary, 2014), echoed by a respondent in Ryan and Garrett who explained that, for them, such searches would be a "breach of confidentiality" (2017, p. 6). Sage and Sage (2016b) note that, from a managerial perspective, such searches on "third party platforms" also mean that confidentiality of service users cannot be guaranteed (Sage & Sage, 2016b).
The literature does report some balancing factors. Clinton et al. (2010) suggest the development of a pragmatic framework that considers the contextual nature of surveillance, outlining six key questions that might be asked prior to conducting a search. These questions covered motivation, threats to the relationship, obtaining informed consent, sharing the found information with the service user, documenting findings and an ongoing need for the practitioner to re-check their motivations for such searches. Other authors were less directive with their guidelines but were firm that guidelines needed to be established (Lehavot, Barnett, & Powers, 2010;Zur, 2012). Most believed that gaining client consent would mitigate against the worst excesses of such searches.

Methodology
In presenting our findings on the use of Facebook by social workers, we are drawing on data from a wider two year study of long-term social work child protection practice (involving fifteen months of fieldwork). The research was ethically approved by the agencies which took part and the research ethics committees of the participating universities. Our key research questions were: How do social workers establish and sustain long term relationships with children and parents in child protection cases? and how are practice, relationships and outcomes for families influenced by organizational cultures, office designs and forms of staff support? Social media use was not a research question for the study.
The fieldwork was conducted in two Local Authorities in England. Researchers were based in the offices of long-term social work teams whose primary role was child protection. Observations were undertaken with the consent of service users/carers and practitioners and managers. Practice encounters between service users and practitioners were observed, detailed notes taken and audio-recorded. The same attention was given to shadowing and recording encounters between staff in their offices, in supervision, observing social workers at their desks and using their computers and mobile devices, and to recording conversations during case conferences.
A fifteen-month fieldwork timeframe enabled us to study up to a full year of casework to draw out the processes of change, or its absence, as it occurred with the aim to shadow thirty child protection cases over the course of one year (fifteen at each site). We felt this number could provide a broad range of cases and enable the depth of inquiry needed. We also aimed to interview service users in the same cases at up to three points during the research period. Our intention was to move beyond one-off observations of encounters to ones that allowed us to shadow long-term practice. We adopted mobile methods that enabled us to shadow social workers everywhere they went, in cars, on foot, around offices, and into and around service user's homes. This approach permitted us to get close to practice to see, hear and feel things (Pink, 2015) that conventional methods like (static) interviews may miss (Ferguson, 2016;Ingold, 2011, Jenson, Sheller, & Wind, 2015. A longitudinal approach also allowed us to observe the mobile, embodied, sensory and emotional dimensions of everyday lived experiences and how social work relationships (or lack of) could impact on the processes of change/no change over time (Neale, 2012;Saldana, 2003). Therefore, our approach (Ferguson et al., 2019) enabled these child protection practices for the first time to be observed in situ over a lengthy period. Our data shows there are real gains in observing what is actually done rather than relying on accounts given. The use of social media had not occurred to us when designing the research and it is unlikely its use would have become evident without us being present over a lengthy period. While the case examples used in what follows reflect actual events and findings, details have been changed to protect the anonymity of the families, professionals and the sites. We were committed from the outset to maintaining the anonymity of the sites where the research took place and no one other than the researchers know the locations and these will never be disclosed.
The detailed field-notes of observations of office routines, formal and informal supervision sessions, conversations about cases, interviews with social workers before and after home visits, observations of practice encounters, interviews with families and transcribed audio records provided a rich source of observational and interview data that was inputted into QSR NVivo 11 software. This tool was used to code the data and conduct thematic analysis and standard techniques of constant comparison were used to produce findings from the data (Bryman, 2012). Files were merged regularly and then re-distributed to the team. This allowed us to collaborate and access the on-going coded work of other team members. A theme to emerge from the analysis of the data was the use of Facebook in ongoing casework, risk assessments and the dilemmas it posed. Once this theme was identified in the main coded data, a further text query search in NVivo was conducted across all collated sources using the terms "Facebook" and "Social Media" allowing us to export a 40,771-word document for further coding and analysis.

Research ethics
During the two-year research period we were not aware of English legislation that proscribed in some situations the use of social media by professionals, namely the Regulation of Investigatory Powers Act (2000) (RIPA) (Directed Surveillance and Covert Human Intelligence Sources Amendment, Order 2010). Under this act the viewing of social networking sites by members of the state (such as social workers) in an investigatory capacity once is not classed as surveillance and does not require RIPA authorization. However, repeatedly viewing social networking sites of service users without their consent, for example to check on their relationships, falls under the definition of surveillance and would appear to require the prior consent of the person posting online. We became aware of RIPA after the analysis and first draft of this paper and we soon learned that our lack of awareness of it was not unusual. Indeed, there are a number of contested interpretations of how RIPA applies to the field of social work and child protection -from instances of judges telling social workers to use Facebook to track down missing parents (Stevenson, 2017), to solicitors and barristers debating its practical implications (Carson & Stevenson, 2017;Community Care, 2018;Reed, 2019;Suesspicious Minds, 2017). Our unfamiliarity with RIPA while undertaking fieldwork meant that we did not ask if social workers had sought legal advice about obtaining RIPA authorization to undertake some of the activities we report; some may have been advised by their legal departments that this was permissible and others may not have sought permission, we do not know. Also, because the use of Facebook was not an explicit research question for the study and we did not anticipate finding it, like the range of unknown and unanticipated things that tend to arise in ethnographic research, it was not necessary to refer to social media use in the ethics application and approval we gained.
Knowing what we knew at the timethat the viewing of service users' Facebook pages was regarded in the codes of ethics as a highly complex and context specific activity, dilemmas that are recognized by the literature (see above for example, Sage & Sage, 2016b;Voshel & Wesala, 2015 ) -we took a decision not to go back to the ethics committee to gain their view on what we were seeing. Our attention was on the primary research questions and the myriad of issues arising about the child protection activities we were studying. Of vital importance was that on the sporadic occasions that some social workers told us they were viewing Facebook pages, they tended to justify these behaviors by articulating how their actions provided valuable information that helped them to keep children safe. To try and stop them looking at Facebook was an ethically complex conundrum because while we were very conscious of the rights of the parents, ethically we had to consider would we have been interfering with social workers' attempts to keep children safe? In trying to understand the culture of social work practices we did what ethnographers are meant to do, observe and elaborate theory to produce the types of findings presented here. During the 15 months of fieldwork, when we became aware of the use of Facebook from time to time at the two sites over numerous cases, we were very careful never to actively participate in social workers use of it. Once we became aware, long after the fieldwork was over, that our data involved potential illegality under RIPA, we sought legal advice from the lead University and guidance regarding ethics from the Chair of the University Ethics Committee. The University takes research ethics seriously and much time was spent exploring and learning from the complex ethical and legal issues involved. The view was that Facebook use was not a research question for the study and so was incidental to it and difficult to predict. Moreover, our discovery of it did not involve us in changing the focus or methods of the research and therefore, there was no requirement under the University's research code for us to apply to the ethics committee for a variation in the original research design and approved ethics for the study. This supported the choice we made to deal with the complex ethical dilemmas in the field as best we could. It should be emphasized that had we known about RIPA at the time of the fieldwork, we would certainly have gone back to the ethics committee because it raised the possibility of illegal activity and we would have addressed the issue with the social workers and managers and asked them if they were aware of the implications of the legislation.
Whilst under RIPA some practices outlined here may be illegal in England, the legal issues are far from clear in other jurisdictions. Regulations, like governments, can come and go but the ethical and power dynamics of child protection work across the globe are enduring. This is why this paper, while taking very seriously issues of legality, uses an ethical frame to present its findings and analysis. The aim is to make a contribution that encourages social work internationally to critically reflect on the ethics of social media use and to engage in debates about what constitutes acceptable/unacceptable behaviors in this highly complex and constantly evolving domain.

Facebook in practice
We have taken quotes from observation notes and interviews with a wide range of respondents to present our findings on the following complex positions taken by social workers: "monitoring", "surveillance" and "drawn in". To set the context, a key finding was that Facebook use was not unusual and practitioners and managers made mention of "office" or "team" Facebook accounts. These were not what are colloquially referred to as "Fake Facebook" accounts, wherein practitioners "befriend" families under the pretense of being someone else, but general local authority Facebook pages that are used to then surveil the Facebook accounts of families' whose setting are not made private. "We've got a team one, we get a lot of information from Facebook", and trails off by saying in a lower voice: "I know it's a bit naughty. … " (SW Manager, observation notes.) These accounts tended to be used to gain information about service users' behaviors, current/past relationships, etc. The following was typical of how we observed this in practice: … (Social Worker) then uses the staff Facebook profile to find [mother's name] and see if there is any evidence of her being in a relationship or using drugs, she finds evidence of both.
[Administrator] helps her to copy the status about her relationship and a photo of what appears to be a bag of cannabis. (Observation notes.) Within the wider network of agencies an awareness of these acts of covert surveillance was sometimes evident. For instance, in a child protection conference: The chair notes that she [mother] is unaware that her Facebook account is being observed, and they wish for this to continue so don't want her to know it is being watched. (Observation notes.) The following social worker talks about their experience of the use of fake Facebook accounts in other teams they've worked in: [Social Worker] reveals that her previous office had a fake profile on Facebook exactly for these purposesto observe the online behavior of service users and "help paint a picture" of them. (Observation notes.) It was possible to conduct surveillance behind the privacy setting set by service users by using fake accounts, however more commonly the service user's lack of privacy settings enabled easy access to their information. Facebook surveillance played a significant role in some cases, for instance in the 12 months of casework we observed with a mother and father who were separated due to his domestic abuse and who was deemed a high risk to their three children. At a case conference five months after we began shadowing, the conference chair asked if there had been "any more Facebook antics?"by which he meant by the parents, not the social worker's use. The social worker indicated that he was monitoring the account and the following exchange occurred between him and a social worker (SW1) from another agency who was involved in the case: [SW1] Her Facebook suggests she has been in some kind of a relationship.
[SW]: Do you spy on Facebook? The use of Facebook is normalized by the chair's easy reference to it which suggests he comes across it regularly at conferences and by the manager's apparent tolerance, if not active support, of it. At the same time, it is also seen as questionable, as "spying", done without the families' knowledge. Six months later due to numerous compelling concerns the children were removed from this family and a social worker was observed telling their manager that they checked Facebook to try and gauge how the mother was feeling. A key finding is that the process of "normalizing" these behaviors enabled social workers not only to justify such actions, but also use Facebook as part of their repertoire of everyday tools for working with families. Social workers who actively participated tended to justify their actions using primarily arguments based on pragmatism, whilst those that refused to use Facebook drew on ethical arguments as explanation. The following section illustrates how social workers actions or inactions were justified.
Facebook was used to see if parents were, in fact, still in relationships with "ex" partners where this contact might pose a risk to children: It can be helpful [to see] if they are in contact with each other … and, to be honest, it's a public domain. So, I mean it is quite intrusive, if you look at it in that way, but if they're putting it up on Facebook, then anyone has access to it. (SW interview.) This quotation illustrates an oft repeated argument made that, because the information posted was publicly available, then it was fair to use it in risk assessments and as part of planning ongoing casework. There seemed to be little or no consideration about rights of privacy or indeed any effort to make the family aware that their Facebook pages were being viewed.
It is quite intrusive, but I feel it's OK if they have no privacy settings and their pages are on public display. They have to gather knowledge from Facebook and take it into account without letting the family know the source. (SW Manager, interview.) This quote again illustrated that, although there was a general acknowledgement that the nature of information gathering was "intrusive", there was still a conscious effort to ensure that the family were kept in the dark about where the evidence came from so as to ensure this potentially rich (and easily accessible) source of information was still available to practitioners. Indeed, some social workers expressed ambivalence over the intrusive nature of social media "monitoring", demonstrating the confusion about the nature of monitoring when covert: … we do a lot of monitoring of Facebook, whether that's a good thing or a bad thing I don't know … (SW interview.) A common argument to justify these actions is encapsulated in the following quote.
… and there was a case where dad wasn't supposed to have any contact with this little girl and there was very clearly a picture on Facebook of dad sat with joint in his hand with baby on his knee. I mean, it's like that thing, isn't it, between what's morally and ethically correct and where, as a local authority, if you're getting evidence from whichever source that tells us that children are at risk we need to be intervening. I don't know where we stand on it, ethically and morally myself I feel like … as a social worker if that's where you're going to get your evidence from … can it be all that bad, if it's going to protect children? (SW interview.) In these cases, practitioners argued that it was justifiable to use surveillance to gather evidence of information already in the public domain. They also tended to argue that to ensure the safety of the child it was necessary to keep the source of information secret from the family. However, when it came to their own online presence: The above quote illustrates that there were contradictions about rights to privacy, as illustrated by this social worker who took a 'practical' position in relation to surveillance of service users.
Social workers who actively took the decision to avoid using Facebook to gather information about families noted variably a lack of skill in social media for not conducting covert surveillance and cited principles such as respect for service users' privacy: Personally, I've not gone seeking out information from people's Facebook. I don't go looking them up on Facebook to see what they up to, because I think that that's an invasion of their privacy … (SW interview.) These arguments used a moral frame to justify not using information posted by service users on Facebook: human dignity, respect for privacy, transparency and trust in the working relationship were important principles in guiding their decision-making in this realm.
Yeah, I wouldn't because I don't know if it's appropriate anyway to snoop … I suppose if you want to look at, check something out maybe, but I wouldn't use it, I've never used it, and I don't know how … . moral it is to, to use it, I don't know really. (SW interview.) A degree of critical thinking was also evident with social workers who questioned the validity of the information posted online, echoing Boyd (2007Boyd ( , 2011. … you can't guarantee that anything that they're putting on there's factual … (SW interview.) As noted above, we found complexity and messiness in the various approaches social workers took to this issue. A "drawn in" theme that emerged from our analysis, which does not appear in any of the current literature, is where social workers were denied their independence to choose whether to view social media posts. This tended to happen when other people sent/presented them with information posted on Facebook.
And people were screenshotting and sending me that, you know … I've had that with quite a few families wanting to show me stuff that different people have put on Facebook. So … families use Facebook as a means of evidencing stuff [so] you are drawn into social media no matter its, even if you choose to not access it yourself, you are drawn into it. (SW interview.) This quote displays a dilemma faced by many social workers, particularly those more likely to think about the complex ethical issues involved in the use of Facebook in child protection work. Personal morals and professional ethics in relation to concerns about respect for privacy, transparency in the working relationship, trust, consent, and reliability of the information posted became secondary when information was sent/presented to them. In these situations, a risk-averse approach tended to trump any actions guided by moral or ethical principles. Our research illustrates that this situation could also be complicated by the following: No, I don't, I haven't got time … I know my manager does a lot of Facebook stalking, loads. He'll tell me things on people's Facebook's about my cases but I, I don't really. (SW interview.) We noted that when a manager used covert Facebook surveillance, a social worker could soon find themselves having to choose whether to act on information provided or risk conflict with their manager. The following quote encapsulated for us the dilemmas that many of our participants faced in relation to engaging with Facebook in child protection: I would love some guidance from somebody to tell us that it's OK to [look at social media] but if you get in evidence like that you've got to, as a social worker, protect that child and obviously safeguarding comes above anything else. (SW interview.) In situations such as these even social workers who would not go seeking this information themselves found that if it came via a third party and they: … see a picture evidence of a child that shouldn't be with a sex offender … (SW interview.) They felt that they would have to challenge the family to explain the picture without stating where they had obtained the information. However, there is a lack of confidence because they can: … think it is sticky ground, so I think it'd be good for the powers above to give social workers some kind of guidance on what they should be doing. Because we know that when you get into court … when these things become unstuck, when you're questioned on information and … how you can validate things, do you know what I mean? … (SW interview.) Some social workers argued that they needed to protect themselves and not be reprimanded for acting on, or not acting on information provided via Facebook. What was apparent was that social workers felt that clear guidance was lacking that could help guide their practices in this area.

Discussion
Our findings confirm the existence of what Sage et al. (2017) described as a middle ground. Although technology is being used, there is a significant level of confusion amongst the organizations, managers and social work practitioners in its usage, all of which is underpinned by an element of risk. In the literature reviewed, some practitioners and organizations were trying to figure out how to get to the next stage, where the use of social media is governed through good practice and policy. As we have shown, our research revealed a lack guidance and confusion about policy, leaving practitioners and managers uncertain about the legality of using Facebook and furthermore the veracity of the "evidence" collected. Participants did demonstrate that if a supervisor approves of social media surveillance, or peers are observed using it, then an individual social worker is more likely to have done so, echoing the findings of Sage et al. (2017) where managers' own practices influenced social worker behavior. In addition, our findings identify an additional, unintentional exposure to social media usage, where service users and their families/associates, or other professionals "draw in" the social worker to social media by pointing out or sending information these others feel is apposite to social work decisions about children's safety. In spite of these factors, activity is still determined by an individual's comfort level with social media. Our findings confirm those of Ryan and Garrett (2017) that some child protection workers were willing to stretch and break ethical codes to use social media, where they felt it was justified in the interests of the child. This practice often focused on checking out whether service users were truthful about their lives and looking for evidence of incongruence (Clary, 2014). Information was used if it came to the social worker accidentally or from a third party, even though there was a lack of clarity about the reliability of the information. For, when considering the kinds of information gained from Facebook, there has been an emphasis on its pragmatic use: "Why wouldn't you use it, when the information might be crucial in protecting children?" is a common refrain. There is a great deal at stake here as information from Facebook seemed to have some influence on aspects of how some families were worked with and could play a part in decisions to remove children from parents. At the same time, practitioner comments revealed a sense of ill-at-ease accommodation of this pragmatic approach. Whether or not there was broad policy or specific legal guidance available on the use of social media, individual social workers tended to weigh up the potential harms and potential benefits in most instances engaging in active dialog of the ethics of their approach with themselves, with supervisors and peers. However, as can be observed from our findings, there were times when ethical compromises took place because the social workers were "drawn into" actions based on other people's social media use. Above all, what we are arguing, is that the ambiguity within the guidance, and failure to keep up with technological advances, has allowed for this to happen, demonstrated by the consequential complex and 'messy' current practice situations we report.

Conclusion
Much greater clarity and guidance is needed for social work leaders, managers and practitioners to help them navigate their way through this digital mine field. As we discovered, there is uncertainty even in the legal profession in England about what is permissible use of Facebook within the law (Reed, 2019). This is particularly pertinent at a time when challenging debates are taking place exploring how social media surveillance is being used at a societal level to monitor, predict and shape the political (Gleibs, 2015) and consumer behaviors (Zuboff, 2019) of users without their explicit consent or awareness. To avoid social work going down a similar morally indefensible road, or more accurately to remove it from the road it is already on, the profession needs to protect service users from unthinking, unethical, and potentially illegal social media use, while opening up discussion around the use of social media as a possible resource in child protection. As the use of social media appears to be driven in practice by managers' participation and role modeling (Sage et al., 2017), to prevent policy and theoretical understandings remaining remote from day to day social work it should be developed in collaboration with these users. This is a sentiment echoed by Mearns et al. (2015) who note that, in the public sector there is a lack of leadership and strategy for working with digital technology. Tregeagle and Darcy (2007) note that technology in social services is dominated by managerialist discourses rather than a willingness to help frontline practitioners and clients to navigate complex decision-making. Our findings suggest that there may always be occasions in high risk child protection work when some social workers feel it is appropriate to check service users' social media accounts. Some might argue this is a defensible practice where the service users' privacy settings permit open viewing of the material. The deliberate deceit of using fake accounts to gain unchecked access to people's lives is unacceptable because of the infringement of legal rights in some jurisdictions, and in all countries the dangers of harm it represents to service users and to the moral worth and integrity of social work as a profession. The rights of practitioners to not indulge in social media or to be drawn in by others, especially managers, also need to be protected. Our findings also have important implications in relation to pre and post-qualifying education in preparing social workers and managers to operate in a socially networked society. These issues are being explored (Boddy & Dominelli, 2017;Taylor, 2017) and innovative teaching and learning social media techniques have been developed in social work education (Cooner, 2014;Westwood, 2014), including a Social Work and Social Media App that enables learning about ethical issues (Cooner, Knowles, & Stout, 2016).
Despite these developments and the existence of legislation and codes of practice, why has our research revealed that covert Facebook use is so common in practice with families? Evidence from our long-term ethnographic approach suggests that it is not just a simple case of "teaching" these requirements to enable workers to apply them. Educational and agency approaches must also consider the environments in which workers operate because, as we have demonstrated, lack of resources, power, uncertainty and risk are central to shaping how social workers engage with the ethics of social media use in a professional capacity (Cooner & Beddoe, 2018). Some of the workers in our study had a strong emotional attachment to Facebook. Through covert surveillance, especially of families who did not want social work involvement and were viewed as uncooperative and secretive, social workers' perusal of Facebook was an attempt to find out more about what was going on in their lives. They justified this practice on a pragmatic basis. It was part of a strategy to cope with anxiety and risk and for workers and the organization to defend themselves against the unbearable feelings of uncertainty about the safety of children (Ferguson, 2018). The culture of social work and child protection that produces these anxieties is unlikely to change unless the emotional complexity of the work and the risks and dangers it deals with, day in, day out is acknowledged, and adequate resources and meaningful emotional support and supervision is provided for staff. Our recommendation, accordingly, is that the current troubling use of social media to covertly monitor families and keep under surveillance those aspects of their lives they choose to share on platforms like Facebook requires the profession to pause and engage with a spectrum of stakeholders in a full discussion to develop guidelines and approaches to staff support, professional development and practice that are relevant to social work in a fast changing digitally networked society.

Disclosure statement
The Authors declare that there is no conflict of interest.