Critical Social Media Analysis: Problematising Online Policy Representations of the Impact of Imprisonment on Families

Abstract

Drawing on the Foucauldian policy analysis framework developed by Bacchi (2009) and building on insights distilled from a study of discourses on the microblogging SNS, Twitter, this paper makes three novel contributions. It unravels how the impact of imprisonment on families is represented in or produced through policy discourses and other governance practices. It also demonstrates how SNS affordances enable affected families to resist and challenge the discourses and proffer alternatives strategies that can inform a transformational problematization model. The paper makes a third contribution by demonstrating how a methodologically innovative triangulation of computational and social science methods can be used to study the contributions of hard-to-reach populations such as the families of people in prison.

Introduction

Online Social Networking Sites (SNSs) constitute important dimensions of the networked, digital forms of “new media” currently transforming traditional modes of communication and information diffusion in the current digital age, including the production and spread of information about criminological and criminal justice issues (e.g., Roberts et al., 2018; Williams et al., 2017). Greer and McLaughlin (2011) for example, contend that social media users are adding their voices and diversifying the meanings and interpretations of key criminal justice discourses. Smith and colleagues (2017, p. 259) go so far as to say that the mediated interactions and communication information emerging from SNSs and other data-driven technologies are, “profoundly transforming social life, identities and relations. . .changing how the social world is configured and how it is engaged with, mediating everyday social practices.”

This paper’s aim is to advance the extant literature on the ways in which SNSs affordances are enabling policy makers and ordinary citizens to frame or reframe discourses about social problems including crime and punishment, with the potential to transform governance cultures as well as cultures of resistance. The concept of affordances in this context refers to the ways in which systems such as SNSs inspire and facilitate particular forms of use, in this context, discourse and knowledge production. MacKenzie et al. (2017, p. 736) note that “the vital quality of affordances is the opportunity for action,” and we argue that discourse is co-constitutive of text (e.g., Tweets) and action (e.g., the act of producing discourses that reframe misrepresentations). Specifically, this paper focuses on how, with the aid of social media affordances, the impact of imprisonment on families is constructed in discourses produced by policy makers and counter discourses produced by affected families. To our knowledge, no other study has explored this topic or analyzed how the families make use of SNS platforms, which as several studies now show, are enabling historically marginalized groups to resist discourses constructed about them and raise awareness of social problems (e.g., Bonilla & Rosa, 2015).

This paper therefore provides new insights that address the dearth of criminological research in this area. It draws on a study that explored the microblogging SNS, Twitter, to analyze discourses produced by policy makers and affected families, about the impact of imprisonment on families. The term policy makers in this context refers to institutions and government officials responsible for prison policy at the national and organizational levels. They comprise the Ministry of Justice, Ministers responsible for prison policy, Her Majesty’s Prison and Probation Services Families Unit, some Members of Parliament, and several prison services.

Whilst we recognize that policies do not translate into frontline practice in any straightforward manner (see e.g., Ugwudike, 2012; Lipsky, 1980), our main focus here is on the ways in which the impact of imprisonment on families is framed in dominant policy discourses espoused by policy makers on SNS sites. The paper is inspired by Bacchi’s (2009, 2012, 2018) Foucauldian, post-structuralist what’s the problem is represented to be (WPR) framework which emphasizes that social policies should be critically interrogated to reveal the representations of social problems embedded in them, which illuminate how governance takes place (see also Bacchi & Goodwin 2016). The framework posits that governance occurs through particular representations of social problems in policy or governance practices and cultures generally, and the representations configure how problems can be understood (Bacchi, 2009). It is from the representations that policy defined “problems” derive ontological reality, and it is on the representations that governance logics, rationalities, and cultures are based. From this perspective, policies do not simply respond to or solve social problems. Instead, social problems are produced through representations inscribed in policies and governance practices. As Bacchi and Goodwin (2016, p. 3) put it: “policies and policy proposals are seen to create or produce “problems” as problems of particular types.”

Using Bacchi’s (2009) framework for critical analysis of policy discourses should therefore involve problematizing policy representations of a social issue. This paper draws on the approach to critically scrutinize policy discourses and reveal how the impact of imprisonment on families is constructed as a particular problematic subject and represented in or produced through policy discourses, reflecting governance practices and cultures. Such critical analysis of policy discourses should unravel the underpinning assumptions and effects of the particular way in which the problem is represented and governed. The WPR approach is additionally useful for investigating how policy representations that silence alternatives can be challenged and reconceptualized. In criminal justice research, the WPR framework has not been applied extensively although it been used for “problem questioning” by a few studies that have critically analyzed drug policies and the effects of associated governance practices (e.g., Brown & Wincup, 2020).

For the purposes of the WPR model, Bacchi and Goodwin (2016) define policy sources broadly to include not only written documents but also other media through which policy discourse is produced and communicated; SNSs therefore constitute relevant policy sources. This paper demonstrates how families, and by extension, other less powerful and socially marginal groups who are affected by criminal justice policies, use the affordances of SNSs facilitated by Web 2.0 to challenge dominant policy discourses and produce alternatives that can broaden understanding and knowledge of social problems and appropriate remedial strategies. Another contribution of the paper is that it also demonstrates how a methodologically innovative triangulation of computational and social science methods can be used to access and study the typically marginalized contributions of hard-to-reach populations such as families affected by imprisonment.

Knowledge Production on Online Social Networking Sites

Digital communication platforms such as Twitter are increasingly evolving into spaces where policy discourses about social issues are produced by policy makers including government institutions who deploy such technologies strategically to communicate their agenda (Bennett, 2018). The sites are therefore facilitating the emergence of new governance practices and cultures, and have evolved into sites of discursive interaction between governments and citizens, in some cases, reflecting open governance and transparency ideals (Bonsón, 2012). Studies do however suggest that several national governments use SNS technologies primarily for disseminating information rather than facilitating democratic eParticipation through citizen involvement in decisions about public policy and the delivery of public services (Bonsón, 2012; Mickoleit, 2014).

Given their power and status, the discourses produced by powerful groups such as policy makers, can become “substantiating discourses” and evolve into accepted knowledge of social problems, as well as key dimensions of governance (Ugwudike, 2015; Foucault, 1977). Policy makers and government institutions can for example, produce discourses and institute governance practices and cultures that reproduce the discourses, influencing how social problems and solutions are understood. Such discourses can silence alternative perspectives including those produced by less powerful groups (see, Lumsden & Morgan, 2017). This phenomenon is captured by Foucault (1977) in his analysis of the politics of knowledge production which highlights the intersections of power, discourse and knowledge production. Without the perspectives of affected groups, the knowledge produced can minimize and devalue their lived experiences. It can also entrench harmful stereotypes and reproduce societal inequalities.

Studies, however, show that the advent of the networked technologies of Web 2.0, such as SNSs, is instituting novel cultures of resistance, enabling typically marginalized and disempowered groups, for example families affected by imprisonment, to resist misrepresentation, proffer remedial discourses, strive for discursive self-determination and participate in political protests for social change (e.g., Bonilla & Rosa, 2015; Dahlberg, 2011). SNSs such as Twitter have therefore become spaces where socially marginalized groups can raise their profile as legitimate actors, challenge dominant discourses whilst at the same time providing insights into their lived experiences (e.g., Dahlberg, 2011). They can frame, reframe, critique, challenge, and correct stigmatizing discourses using countervailing responses (e.g., Bonilla & Rosa, 2015). They can also help raise awareness and engage in forms of internet-mediated collective action, activism, contestation, and shared global solidarity with diverse actors to prompt social change. Commonly cited examples include forms of digital activism such as the #Black Lives Matter (#BLM) campaign (Bonilla & Rosa, 2015; Carney, 2016) and the #MeToo movement (Clark, 2016).

From this milieu, marginalized groups campaigning for policy reform and social change emerge as “produsers” actively involved in creating a new culture of democratized knowledge production whilst harnessing other communicative and informational benefits of networked ecosystems and their viral properties. It is nevertheless useful to nuance the variegated benefits of interactive web-based technologies with the caveat that usage can be channeled toward socially harmful behaviors. Instances of this phenomenon include online radicalization and extremism (e.g., Awan, 2017, see generally, Jewkes & Yar, 2011). Another example is the propagation of “fake news” as has been the case during the Covid-19 pandemic (Depoux et al., 2020). But as we have seen, other studies reveal that if used for prosocial purposes the sites can provide profound social benefits.

The foregoing discussion points to the proliferation of digital platforms such as the SNSs now providing opportunities for typically marginalized groups to frame or reframe the discourses that affect them. Families affected by imprisonment represent a key example. In England and Wales, there were just under 80,000 in prison in 2020 (World Population Brief, 2020). Most prisoners (over 70%) have families who offer them financial, emotional, and other support but studies reveal that many families experience social, financial, psychological and practical difficulties when a family member is imprisoned (see Condry & Smith, 2018 for a comprehensive review). As a result, families become victims of vicarious punishment (henceforth, VP) as these difficulties represent forms of penalization that stem directly from the punishment imposed on someone else (their family member).

Framing Policy Discourse, Governance, and Counter Discourse

The discussion so far reveals that although the discourses produced on SNSs by powerful groups such as policy makers can exert epistemic domination (see also, Ugwudike, 2020), SNS affordances also empower historically disempowered and marginalized groups such as families affected by vicarious punishment, to produce counter discourses that inform knowledge and understanding of social problems. Drawing on Bacchi’s WPR approach, the paper analyses how the impact of imprisonment on families is represented in policy discourses and how the families of people in prison produce alternative and less harmful conceptualizations of the problem.

The WPR framework stresses that “issues are problematized—constituted as ‘problems’ – within policies” (Bacchi & Goodwin, 2016, p. 3) and problematizations are as such inextricably linked to governance practices. Therefore, policy problems are devoid of immanent properties and should be critically scrutinized to unravel how people and issues are produced as problems and governed, since governance takes place through problematizations (see also, Bacchi, 2009, p. 1; Bacchi & Goodwin, 2016). As such, policies and governance practices are not developed in direct response to social problems. Instead, policies actually produce representations of “problems,” entrenching particular understandings of social problems and foreclosing alternative conceptualizations (Bacchi, 2009, p. 1). As Bacchi and Goodwin (2016, p. 3) note, “policies and policy proposals are seen to create or produce “problems” as problems of particular types.” It is therefore important to investigate representations of social problems in policies. Such analysis is also necessary because policy representations are central to governance practices and cultures that determine the treatment of affected groups (Bacchi, 2009), in this case, families affected by imprisonment.

Methodology

The Stakeholders

The study retrieved publicly available content posted on Twitter by two identified stakeholders: policy makers and families affected by imprisonment. A broad and inclusive definition of families was adopted although most of the retrieved posts were automatically retrieved using a software which is described below. The paper also includes user-led advocacy groups providing support and advice services to prisoners’ families.

Computational Data Collection

Twitter data posted by both stakeholders were extracted within a data capture period of 01/01/2020–01/07/2020. The study used a data extraction software, Web Data Research Assistant (Version 3.5.11) (WebDataRA) (Carr 2020) which is a social media mining tool that uses Twitter’s application programming interface (API), to scrape historical and real time data directly from Twitter. WebDataRA retrieves and archives internet data such as Twitter data through browser automation, and in full compliance with Twitter’s terms and conditions. To preserve the anonymity of users, in line with the institutional ethical approval process for this study, account names and Twitter handles are not included in this paper, and only brief but insightful excerpts from the retrieved data are provided.

The search terms that were used for data extraction were: “prisoners” followed by each of these words: families, family, wife, husband, partner, children, brothers, sisters, parents, mum, dad. As well as each of these words, followed by the phrase “of prisoner/s.” These search terms were chosen because the impact of imprisonment on families has never been a “trending” topic with a prominent “hashtag” connecting key users and central debates about the topic that can be used to locate relevant tweets instead of having to rely on various search terms.

Following data extraction, a data cleaning process was instituted to delete content that did not meet the following criteria for inclusion—(1) posts about prisoners’ families by the two stakeholders, and (2) posts from outside England and Wales. Having removed irrelevant content, the study focused on an in-depth analysis of 20,000 tweets comprising 10,000 tweets from each of the two stakeholders: the families and policy makers.

Computational and manual data analysis

In-depth analysis of the retrieved tweets was conducted using the Corpus Assisted Discourse Studies method (e.g., Baker, 2010) which combines software-driven quantitative analysis of textual data with manual qualitative analysis. Our application of this method involved the following:

• 1) Saving the retrieved tweets as an excel spreadsheet.

• 2) Uploading the corpus of 10,000 tweets per stakeholder to Voyant Tools (2.4) (Sinclair & Geoffrey, 2020) which is an opensource, web-based software that organizes digital data. The software performs statistical analysis of text retrieved from digital communication technologies such as SNSs. It also, amongst other things, quantitatively identifies keywords which are the terms that are topical and significant in the corpus and to the core discourses. Table 1 below sets out the top relevant keywords found in each of the two corpora of tweets uploaded to Voyant Tools 2.4.

Table 1.

Keywords in Each Corpora of Tweets.

Stakeholder	Relevant keyword
Families	Prison
Policy makers	Visits

To broaden the depth and scope of analysis, we used the term (visit*) to conduct a “wildcard” search of the policy makers’ corpus of 10,000 tweets, and the term (fam*) to search the families’ corpus of 10,000 tweets. The aim was to identify and analyse all the occurrences and permutations of each keyword (in Table 1) in the relevant corpus. Placing an asterisk after a search term yields results that show all the terms which start with the prefix (before the asterisk) but end differently. For example, the search term “visit*” within the policy makers’ corpus produced all the words associated with the term in the corpus:

• visit, visits, visitation, visited, visiting, visitor, visitors, visitscentre, and visitsupdate.

The search term (prison*) in the families’ corpus yielded several terms including ¹ :

• prison, prison’s, priosn2020, prisoner’s, prisonerfamily, prisonerlivesmatter, prisonermum, prisonerpeopletoo, prisonersfamilies, prisonersfamiliesmatter, priosnersarelovedones prisonvisits, prisonfamilies, prisonfamiliesvote, and prisonwives.

Having established the relevant keywords with which each stakeholder constructed discourses about VP, we used Bacchi’s (2009) WPR framework to analyse the policy makers’ discourses and identify the representation of the VP problem as well as the prevailing mode of governance that produces the problem. To understand the effects of the dominant problematization, and identify alternative problematizations as well as their corresponding solutions, we analyzed how affected families frame their experiences of VP. Specially, we were guided by key interrelated questions fundamental to the WPR framework:

1. What’s the problem represented to be in a specific policy or policies?

2. What deep-seated presuppositions or assumptions (conceptual logics) underlie this representation of the “problem” (problem problematization)?

3. How has this representation of the “problem” come about?

4. What is left unproblematic in this problem representation? Where are the silences? Can the “problem” be conceptualized differently?

5. What effects (discursive, subjectification, lived) are produced by this representation of the “problem”?

6. How and where has this representation of the “problem” been produced, disseminated, and defended? How has it been and/or how can it be disrupted and replaced?

Step 7: Apply this list of questions to your own problem re-presentations

(See, Bacchi & Goodwin, 2016, p. 20). We used questions 1-6 in the list of questions above to analyse the discourses produced by the policymakers. Further, we assessed whether, and how the family members’ discourses can help address questions 5 and 6, and inform alternative, responsive problematizations and allied solutions.

CDAS: Concordance analysis

Our investigations consisted of concordance analysis which is a Computer Assisted Discourse Studies technique. This was done using Voyant Tools 2.0 which generated a set of concordance lines or a Keyword in Context (KWIC) display showing the keywords (listed earlier) in the original context in which they were used by each stakeholder. The KWIC display was useful for analyzing tweets using the WPR approach without focusing solely on word frequency analysis. A sample KWIC display is provided in Table 2 below which shows the keyword “prison” in the middle column (as used by the families) with the texts surrounding it on the left and right. Each row of texts can be expanded for more text and context.

Table 2.

Sample KWIC Display of the Keyword Prison.

for the best Amen Just because my boyfriend is in	Prison	& I decided to stay by his side doesn’t mean that
than other people’s opinions & the stigma that comes with a	Prison	sentence #prisonerspeopletoo LETS SAVE OUR PRISONERS LETS. . .
your loved one is not a part of going to	Prison	#prisonerspeopletoo ever since visits were cancelled and the lack of
partner is also struggling with the uncertainty of the situation	Prison	is about deprivation of liberty not life. Families deserve answers
not life. Families deserve answers. Having a loved one in	Prison	is hard enough as it is, but not being able
it hasn’t been easy but we can get through anything. #	Prison	#love #mentalhealth #relationships I am proud and so pleased about
time. Being banged up for hours and hours is heartbreaking. #	Prison	#lifer #relationships So pleased we are at least getting credit
few extra pounds for his canteen. It’s the little things. #	Prison	#love #relationships When are you going to talk about prisoners

We are able to reproduce this table here because the tweets are not written out in full; several words are missing. This means that we can preserve the anonymity of the authors.

The KWIC tables enabled manual in-depth qualitative analysis of the keywords within the different contexts in which they were used by the stakeholders. The analysis entailed reading and coding each line of the KWIC (concordance) table in each of the two corpora of tweets. Codes depicting the ways in which each of the two stakeholders used their keywords in their corpus to frame discourses about VP were applied. Once we completed initial coding, we connected and categorized mutual and interrelated codes reflecting the same overarching themes.

Methodological limitations

Social media analysis such as the one provided here, necessarily focuses on a proportion of available content. Consequently, many users including private accounts and non-users cannot be included in the analysis. But an ameliorative factor is that such analytics can reach large groups of stigmatized and typically hidden populations and can generate more substantial data than regularly used social science methods.

Another problem concerns the presence of fake accounts including “software robots” known colloquially as “bots” (Ferrara et al., 2016). This problem did not particularly affect the policy makers’ corpus of tweets because all were either posted by verified accounts or well-known institutions. Regarding the families’ data, we used WebDataRA’s technical enhancements which provide several account authentication techniques. For example, the tool provides metadata that can be used to ascertain the veracity of an account. High frequency posts by accounts with limited connections to others (through “follows” for example) can indicate unreliability (Yang et al., 2019). But in our study, we were mindful that many families could fall within this category. Our verification processes therefore had to involve a painstaking process of examining the content posted by accounts we tagged as suspicious, and this lasted several months.

Results

Producing the Problem of Vicarious Punishment in Policy Discourse

We grouped all the codes that emerged from the KWIC analysis of policy makers’ keywords into the following themes, (1) loss of familial influence and (2) responsibilization. The theme that emerged from the KWIC analysis as useful for addressing the first question in the WPR framework which is: “What’s the problem represented to be in a specific policy or policies”, is loss of familial influence. This constitutes the representation of the problem of VP. In response to the second question within the WPR model, a key assumption underlying this problematization is that loss of familial influence during imprisonment can deprive people in prison of the rehabilitative role of family members, increase risks of reoffending, and undermine public protection. It can also undermine the mental wellbeing of people in prison, with adverse implication for prison order. Thus, the core representation of the problem of VP focuses on the implications for maintaining law and order rather than the harmful impact on affected families coping with the imprisonment of a family member. A second assumption that emerged from the KWIC analysis is that affected families should be responsible for, and capable of, providing the contact and support required for improving the wellbeing of people in prison, to maintain prison order and should be able to support rehabilitation.

What is left unproblematic by this problem representation (see Bacchi, 2009) is the broad range of hardships associated with the VP of affected families. In the policy makers’ representation of the problem there are deafening silences surrounding the multifaceted problems families encounter, foreclosing much-needed consideration of those problems, some of which constitute obstacles to maintaining contact. For example, studies show that women predominantly assume responsibility for providing such support for their partner or child in prison and experience many financial, emotional, and other hardships as a result (Jardine, 2017) including secondary prisonization (Comfort, 2003, 2019).

The concept of secondary prisonization expands on the notion of primary prisonization, which Clemmer (1940) used to explain how prisoners adapt to the pains and deprivations of the prison environment (see also, Crewe, 2011; Sykes, 1958). Comfort (2003) reframed the concept as secondary prisonization, to highlight the experiences of affected families. The concept refers to the adverse practical experiences of female family members who often bear the burden of supporting loved ones in prison and have to undergo dehumanizing interactions with prison personnel. Such interactions include intrusive searches during prison visits. Despite this, the KWIC analysis showed that the prevailing representation of the problem in policy discourses obfuscates these and other harms of vicarious punishment. Instead, it foregrounds the criminal justice imperatives of maintaining prison order and securing rehabilitation.

Responsibilization

The second theme that emerged from the concordance (KWIC) analysis of the policy makers’ corpus of tweets is “responsibilization.”This refers to the policy or governance practice in which the aforementioned problematization (loss of familial influence) is embedded, which also reproduces the problematization. The governance practice involves facilitating family visits and other means of contact (subject to risk assessments) primarily for their role in maintaining prison order and supporting rehabilitation.

Whilst maintaining family contact and ties is very important and should be facilitated by all means, defining such contact as the core “solution” to the problems related to the imprisonment of a family member and justifying it on the basis of criminal justice imperatives, is a responsibilization technique that simultaneously minimizes the problem and devolves responsibility for addressing it to the family. It is therefore a family governance technique that seeks to “responsibilize” families whilst absolving the state of its role or minimizing the need for official intervention. This mode of governance has a neoliberal orientation which penal scholars distinguish from post-war expansionary policies (e.g., Garland, 2001). As developed by Garland (2001), responsibilization is a by-product of the neoliberal agenda of scaling down state involvement in welfare provision and devolving responsibility for crime reduction and other roles traditionally reserved for the state to non-state actors, in this case families.

The responsibilization of families was evident in several posts. For example, there were tweets about “building the importance of family ties,” and tweets that stated that they are “vitally important to prison and prisoners” and “crucial to prisoners” “wellbeing” and that it plays a “vital role in rehabilitation.” Thus, families were depicted as instruments of rehabilitation. Further, alongside governance practices of facilitating visits where families can perform their rehabilitative role, families were invited to engage in other rehabilitative activities in prison such as parenting and family relationship programs, therapeutic activities and events for prisoners, examples of which include creative arts activities and sporting events. These were framed as activities that are expected to bolster the rehabilitative potential of maintaining contact.

In sum, the concordance analysis revealed that, loss of familial influence resulting in adverse implications for prisoners’ wellbeing and rehabilitative goals, was a recurring theme in policy discourses about the problem of VP. It was the main representation of the problem and the proposed “solution” took the form of responsibilization whereby families were expected to assume responsibility for addressing the problem.

Origins of the Representation of the Problem as “Loss of Familial Influence”

To a significant extent “loss of familial influence” is a problem representation that can be understood in the context of broader policy developments. It particularly dovetails the emergence of a neo-liberal penal order that prioritizes technocratic modes of governance as noted by Feeley and Simon (1992) and Garland (2001) in their analysis of the shift in the ethos of criminal justice from transformative principles such as rehabilitation to systemic priorities of cost effectiveness and efficiency. Part of this shift has involved assigning some of the costs of criminal justice to non-state actors, in this case families.

The problematization is also inspired by law-and-order punitiveness and the longstanding quest of successive governments to project a “tough on crime” image (e.g., Jennings et al., 2017), an image that seems at odds with a welfarist, family-centered approach to dealing with the problem of VP. “Courtesy stigma” as conceptualized by Goffman (1963) which manifests itself in the perception that families are guilty by association (see also, Kotova, 2019), can also discourage the enactment of more expansive policies for families, encouraging instead the focus on a reductionist representation of the problem, a representation that foregrounds criminal justice imperatives.

How and Where Has This Representation of the “Problem” Been Produced, Disseminated, and Defended?

According to the WPR model, another question worth considering is how and where the representation of the “problem” has been produced, disseminated, and defended. When applied to the prevailing representation of the VP problem in terms of “loss of familial influence” resulting in adverse implications for prison order and rehabilitation, we find that beyond policy discourses on SNSs, it is a discourse that also features in policy documents. For example, in a recent document, the MOJ in England stated that, “Family networks . . . can contribute to reducing reoffending, and assist resettling into communities” (e.g., Ministry of Justice, 2020).

In England and Wales, a core theme that emerged from a report on the family ties of people in prison, which was commissioned by the Ministry of Justice, is that family contact and ties should be central to prison reform. The report also stated that, “harnessing the resource of good family relationships must be a golden thread running through the processes of all prisons. . .” (Farmer, 2017, p. 4).

Since the report was published, efforts to responsibilize families have intensified. For example, a recent policy framework set out in a document titled Family and Significant Other Strategy and development plan (Ministry of Justice, 2019) made it clear that efficient family-related work is fundamental to “reducing reoffending, keeping prisoners safe and preventing self-harm and suicide.” This new policy framework requires all prison services to establish a Family and Significant Other Strategy and a Family and Significant Other Shadow Measurement Tool created by the Ministry of Justice (2019) for reviewing and scoring the quality of family practice. These requirements were introduced in response to recommendations of the Farmer report and it emphasizes the importance of facilitating “social visits” to support rehabilitation and resettlement, and also address other problems such as “self-harm and suicide.” It places additional responsibility on families by seeking to make them more actively involved in providing the support required for the rehabilitation and resettlement of their family member in prison, and for preventing other serious harms in prison.

Reconceptualizing the Problem

Two additional questions that are central to Bacchi’s (2009) WPR approach are: “Can the problem be conceptualized differently?” and “What effects are produced by this representation of the problem?” To address these questions, it is useful to understand how groups (such as affected families) who are most affected by a social problem frame their experiences. To this end, we analyzed the contexts in which affected families used the top relevant keyword in their corpus of tweets, (“prison”) as well as the associated terms set out earlier, to describe their experiences of VP and propose solutions. We grouped all the codes that emerged from the KWIC analysis of these terms into an overarching theme which is The Pains of Vicarious Punishment.

The Pains of Vicarious Punishment and the Effects of the Problem Representation

As a core theme that emerged from the analysis of the families’ corpus, The Pains of Vicarious Punishment illuminates the diversity of ways that the families defined the impact of imprisonment and also shows how the problem can be reconceptualized. Sykes (1958) originally used the concept of “pains” to describe the pains and deprivations of the prison environment (see also, Crewe, 2011). Here we expand the concept to reflect the adverse impact of imprisonment on the family members of people in prison.

Explaining Lived Effects of the Prevailing Problematization and Producing Alternatives

In describing the Pains of Vicarious Punishment, the families highlighted specific problems. These constitute key “lived effects” of the narrowly constituted ontological representation of the problem as “loss of familial influence” and the allied governance practice of responsibilization. In the focus on loss of familial contact, the harms of VP are ignored or minimized. Bacchi’s WPR model calls for an analysis of such “lived effects” of representing and governing a social problem in a certain way and these are explored below. Again, the terms in inverted commas are in vivo codes (terms from actual tweets).

The impact of separation

A lived effect of the policy representation of loss of familial influence as the core problem is that it ignores the impact of separation caused by the imprisonment of a family member. The difficulties associated with this were emphasized by most of the affected families. There were references to the emotional impact of separation on children and descriptions of developmental milestones missed by the parent in prison. Examples were a child’s “first words” and missed birthdays. The emotional impact of separation was also described using terms such as “trauma,” “heartbreak,” “grief,” “unforeseen tragedy,” “sleeplessness” “struggle,” and looking down a “dark tunnel.” Some of these emotional pains of VP became more acute immediately after each visit to the family member in prison. Studies similarly reveal that affected family members experience psychological difficulties primarily stemming from the emotional stress of separation from a family member, and the disruption of family life and relationships (e.g., see Author, 2019).

In part, the emotional impact of separation was exacerbated by their concern for the family member in prison experiencing the pains and deprivations of imprisonment. This concern became more pressing during the pandemic and the subsequent imposition of severely restrictive conditions in prisons where prisoners became mostly confined to their cells. The imposition of severe “lockdown” restrictions as well as banned visits was confirmed in a Ministry of Justice (2020) document and condemned by human rights organizations and others (e.g., Howard League for Penal Reform, 2020). Most family members expressed serious concerns about the impact on the mental health of loved ones held in such austere conditions, which they described as inhumane, stating that the people in prison were being held like “battery farm animals” and prisons had become “Covid 19 death traps.” They felt that the practice undermined the duty of care imposed on the prison service “to protect prisoners.” They also highlighted how easily the rights of people in prison and their families can be abandoned without consequence. There were references to “human rights being broken for prisoners and their families” and “human rights being breached” as well as the view that the austere conditions are “against human rights & right now, prisoners & their families have been forgotten.”

Barriers to contact represent another dimension of the adverse impact of separation as described by the families, with some family members decrying obstacles such as the costs of maintaining contact including what they perceived to be exorbitant telephone charges. In their view, this amounted to profiteering and deriving financial benefits from the “trauma” of others. Here we witness another pain of VP that is masked by the reductionist policy representation of the impact of imprisonment on families. It pertains directly to the material implications of VP which exacerbates the existing socioeconomic marginalization of affected families. A fundamental theme arising from studies of imprisonment in Western jurisdictions with a large prison population, such as England and Wales, and the US, is that many prisoners and their families experience socioeconomic marginalization and social exclusion (e.g., Jardine, 2017, 2018; Ministry of Justice, 2012). These problems can become more intense when a family member is imprisoned. For instance, financial difficulties can become more severe where the prisoner was the breadwinner or where the family loses the prisoner’s social security benefits. Affected families will struggle to sustain the costs of maintaining contact through costly telephone calls and prison visits (see generally, Condry & Smith, 2018).

Perceived injustice

For some family members, a sense of injustice adds another dimension to the pains of vicarious punishment, and it constitutes yet another lived effect of the narrow policy representation which obscures the pains and the need for responsive solutions. The sense of injustice manifested itself in the belief that families were not being accorded the considerations, rights, and privileges that others enjoy. For example, as initial lockdown restrictions were lifted, some families wondered why people could visit recreational venues such as cinemas and zoos, but families could not visit family members in prison even after they had been separated for months because of prison “lockdown” restrictions.

The perception that families are subject to inequitable treatment was attributed by some family members to the stigmatized status ascribed to them, a by-product of “courtesy stigma” (Goffman, 1963). Indeed, the KWIC analysis also revealed how some families experienced this stigma, with some referring to the loss of friends over time. Some also felt that the stigmatization of families breeds lack of compassion amongst the public and contributes to poor treatment in prison during visits, which according to Comfort (2003) is a dimension of secondary prisonization.

Another aspect of this problem is the negative impact on the self-identity of affected families. Reflecting this, a particular concern pertained to intrusive searches which were perceived by some family members to be a manifestation of “courtesy stigma.” In their view, this form of stigmatization fuels the suspicion amongst prison services that families are more criminogenic than others and as such more likely to transport contraband items such as drugs into the prison. Such stereotyping and stigmatization can foment a negative self-identity, as also noted by Comfort (2003) in her study of secondary prisonization.

Lack of recognition

Another core dimension of the pains of VP is Lack of recognition by policy makers. It is another lived effect of the reductionist policy representation. There was a sense amongst family members that they were viewed by the policy makers as “nobodies,” and were “disregarded” and that there was a longstanding tendency to “treat family members of prisoners as a nuisance.” These and several terms also revealed the impact of the perceived lack of recognition on the families’ self-identity and esteem.

This perception found expression in the many comments about lack of adequate information from policy makers regarding prison policies, and “being ignored” particularly by key government ministers and also by institutions such as some prison services. This again demonstrates how dimensions of secondary prisonization as described by Comfort (2003) migrate from the social world to the digital public sphere and re-emerge as tertiary prisonization whereby families become susceptible to institutional disregard, undermining their experiences and self-identity. The view expressed by a family member that; families are frequently “treated as bad as prisoners” encapsulates this point.

Discursive and Subjectification Effects of the Prevailing Problematization

Added to the “lived effects” of the policy representation that ignores or minimizes the pains of vicarious punishment, the KWIC analysis of the families’ corpus revealed another important effect which according to the WPR model can be described as a “discursive effect.” The effect is conceptualized here as “digitized epistemic domination” (see also, Author, 2020) and it is the product of the prevailing policy representation which entrenches the belief that loss of familial influence is the key problem and pays insufficient attention to the pains of vicarious punishment. The discourse dominates policy documents as already noted.

The ways in which social problems are represented in policy and governed can also produce what Bacchi’s (2009) WPR framework conceptualizes as a “subjectification effect.”

This effect manifests itself when representations and governance produce social relationships in which groups are differentiated according to certain criteria that disadvantage some and benefit others. For example, representing the VP problem in terms of loss of familial influence required for prison order and rehabilitation, and instituting responsibilization strategies such as prison visits during which parenting and family programs can be delivered to enhance the rehabilitative potential of families, as was in the case in some prison services included in the study, can trigger the subjectification of families who appear unwilling or unable to fulfill their ascribed rehabilitative role. This could involve ignoring the myriad of problems that can explain their absence and labeling such families as the problematic object that is deficient in some way compared with other families.

This subjectification process underscores Bacchi and Goodwin’s (2016) position that problematizations and governance practices embody assumptions about those at whom the governance practices are targeted. It is evident in several family policies such as those that seek to regulate so-called “troubled families” (Crossley, 2018), or “failing families” (Pitts, 2001, p. 97) through governance practices that aim to resocialize and transform them into the ideal neoliberal family that is self-sufficient with an aptitude for self-governance or self-regulation.

Reinforcing this, the aforementioned report by Farmer (2017, p. 4) on families’ contact and ties during imprisonment, stated that, “Families need to be willing and able to engage with the rehabilitation process.” Indeed, it is possible that the effect of “courtesy stigma” is such that the families of people in prison are generally viewed at policy level as lacking the qualities of the ideal neoliberal family, hence the provision of responsibilization strategies such as parenting and family relationship programs in several prisons, to “re-socialize” the families. Families considered deficient in neoliberal family values have long been singled out for stigmatizing discourses and governance techniques that seek to responsibilize and transform them into productive neoliberal citizens (see e.g., Rose & Miller, 2010).

Responsibilization techniques in such cases include interventions such as parenting skills and family relationship programs (e.g., Social Mobility Commission, 2017) which as the KWIC analysis revealed, were part of the package of interventions offered to in prison to residents and visiting families. These governance practices which seek to regulate families also produce and reproduce the reductionist representation of the problem. In doing so, they reinforce the neoliberal minimization of societal inequalities that can shape family structures and experiences. They can therefore legitimize family-level state intervention and minimize the need for societal-level reforms.

Discussion and Conclusion: Disrupting and Replacing the Representation of the Problem

In this section, we consider a final question within the WPR approach which is this: how can the representation of the problem in policy be disrupted and replaced? With this question, alternatives to the problematization and governance of social issues can be explored.

In addressing the question, we propose a transformational problematization model that is responsive to the lived realities of affected families as defined by them. Central to this model is the view that any representation of the problem should embody the perspectives of those most affected by the problem, in this case the affected families. Therefore, the proposed transformational model emerged from the KWIC analysis of the families’ corpus of tweets.

A key dimension of the model is that its representation of the problem focuses on “The Pains of Vicarious Punishment” outlined earlier. These are expansive in their deep and broad elaboration of the impact of imprisonment on families unlike the reductionist policy of problematization. Equally embedded in this new model are corresponding solutions which are again rooted in themes that emerged from KWIC analysis of the families’ corpus. The themes are: Representation and Amplification, and they are discussed below. In focusing on solutions proposed by the families, this paper adopts the normative ethos of Bacchi’s (2009) WPR framework which highlights the need to challenge policy representations and “take the side of those who are harmed.”

Representation

The families’ corpus of tweets comprised references to their perceived position as a group with limited representation and power to formulate discourses that would garner policy recognition of the pains of vicarious punishment. There were calls for representation at an institutional level, for example, in the form of a “prisoners” families commissioner. In one of the comments alluding to this, a family member said, “we need a prisoner’s families commissioner, one of us, who will stand up for us. We have been ignored for too long.” In this paper, we similarly argue that an Independent Commissioner is indeed needed and should be statutorily required to protect the rights of affected families and provide requisite advocacy.

Part of this could include establishing frameworks for holding policy makers accountable for the welfare of affected families. At present, families lack requisite representation; there is no independent, officially recognized body representing them. Such official bodies have been rightly established to support other groups. Examples include the Victims’ Commissioner for England and Wales and the Children’s Commissioner for England. In the case of families who are victims of vicarious punishment, Third Sector Organizations currently dominate the provision of support services. A “families group” has also been established within Her Majesty’s Prison and Probation Services. This group produces operational guides and policy frameworks for supporting prisoners to maintain relationships with families on the basis that such relationships prevent reoffending (see, HMPPS Families Group, 2017). Nevertheless, the lack of independent representation should form part of any problematization of the impact of imprisonment on families, not least because such a problematization would inspire the introduction of measures for ensuring that the rights of affected families and pains of VP are recognized at policy level and responsive measures are instituted.

Amplification

The KWIC analysis also found many references to the importance of striving for online visibility to raise awareness of the pains of vicarious punishment. Some of the fundamental pains identified in the corpus of tweets have already been outlined in previous sections. Meanwhile, the efforts of affected families to raise awareness of the pains was evident in their numerous tweets and the campaigns some of them instituted to directly communicate their adverse experiences as well as the harms of imprisonment to all and to pressurize policy makers to introduce responsive polices. They called for others to “speak up” and maintained that they were “fighting for justice.”

Insights from the families’ awareness raising campaigns could be harnessed by policy makers to develop a responsive problematization of the problem. As noted earlier, we recognize that centralized policies do not translate straightforwardly into organizational policy or even frontline practice. Lipsky (1980) notes in his seminal work on the policy making role of frontline public service workers that their discretionary decisions and practices culminate into “agency policy.” This has been demonstrated in several institutional contexts such as policing Devroe et al. (2017) and probation practice (e.g., Author, 2011, 2012). Indeed, some argue that “penality-in practice” (that is, the impact of penal discourses on actual frontline practice) can be unexpectedly distinct from penal policy provisions (Author, 2012; Cheliotis, 2006). But the WPR alerts us to the fact that there can be a dominant policy discourse that shapes how governance happens. In the current study, the dominant discourse appeared to permeate the different levels of policy making and practice, from central government to the organizational and frontline (prison service) level.

To include the families’ perspectives and rebalance the dominant discourse, policy makers at all levels could pay attention to the pains of VP as articulated by the affected families themselves. Part of this can involve engaging with families and the user groups representing them across several arenas including SNSs. Regarding SNSs specifically, there were numerous references in the family members’ corpus to the lack of response from key policy makers, primarily from those at the center (e.g., government ministers) but also from those operating at the organizational and frontline level, namely some prison services. This was the case, despite the families’ efforts to communicate with them. The view that families are “invisible” and “not heard,” and that their Tweets should at least be acknowledged “to show support” was a recurring theme. For example, one family member directed this question at the Ministry of Justice, a government Minister and a Member of Parliament (using Twitter mentions), “Why are the feelings of families being ignored? Why is ok to care about everyone else but not us? Prisoners’ families MATTER.”

A key benefit of responding to the families on SNSs is that if fosters interactions between citizens and policy makers on such sites. Such connections can amplify the voices of those most effected by a social problem, enhancing the possibility that their voices will influence policy decisions. In this respect, the affordances of SNSs cannot be over-estimated. As noted earlier, studies show that SNSs such as Twitter are increasingly transforming the landscape of knowledge production, enabling diverse voices to participate in the framing of social issues, and empowering them to challenge dominant narratives through various forms of digital activism (e.g., Bonilla & Rosa, 2015; Carney 2016; Freelon et al., 2016). In a critical analysis of digital media and democratic participation, Dahlberg (2011, p. 861) also notes that digital spaces can enable online “discursive contestation” in the form of “contestations of power. . .strengthening the voice of alternative, marginalized, or otherwise oppressed groups.”

Together, these studies suggest that social media spaces can enable traditionally marginalized groups to develop counter narratives and resist epistemic dominance which can entrench stigmatizing stereotypes and reproduce inequalities by fueling social isolation and exclusion. Policy makers and public authorities engaging with SNSs could draw on insights produced by affected families on those sites, to develop responsive family-centered measures. Indeed, law enforcement agencies and even politicians as well as others capable of influencing criminal justice policy are increasingly relying on the knowledge that is produced and disseminated across data driven SNSs, about crime and criminal justice (see, Walsh & O’Connor, 2019).

In sum, a problematization model that emphasizes lack of representation and the need for amplification should produce effects that are likely to be beneficial to families. First, it can give rise to remedial lived effects such as the amelioration of the pains of VP via policy recognition and responsive measures. Second, it can produce discursive effects that go some way toward reversing the dominant problematization (loss of familial influence) which is currently bolstered by the neo-liberal governance technique of responsibilization, a technique that minimizes the state’s role in welfare provision. It can as such help entrench the notion that policy makers, not families, are the fitting targets of responsibilization. Third, it can address subjectification effects whereby affected families are portrayed as deserving of sometimes punitive and intrusive responsibilization techniques that seek to produce the self-sufficient and self-regulating neoliberal citizen. The techniques ignore societal inequalities that are exacerbated by VP and which therefore require state intervention and support.