Regulating Front-Line Workers: Rules, Review, and Roles

Welfare programs have long regulated both the distributers and recipients of public benefits, yet scholars tend to examine welfare’s client-disciplining effects. Studies of early welfare programs typically consider how caseworkers disciplined people living in poverty by differentiating between the deserving and undeserving poor, but overlook how these programs regulated caseworkers themselves (Piven and Cloward Reference Piven and Cloward1993; Handler and Hollingsworth Reference Handler and Jane Hollingsworth1971). For instance, research on single mothers’ allowances in Canada identifies how a small set of broadly worded rules enabled caseworkers to use personal qualities, such as marital status, British subjectship, or being a “fit and proper” mother, to distinguish those who were deserving of state-funded assistance from those who were not (Gavigan and Chunn Reference Gavigan, Chunn, Gavigan and Chunn2010, 53–57). These early programs relied on caseworkers to skilfully assess the homes and personal circumstances of sole-support mothers and provide some (but not others) with individually-tailored supports. By allowing caseworkers to penalize or reward individuals based on how closely they fit the normative model of a deserving welfare recipient, these programs used caseworkers’ discretion as a tool to regulate sole-support mothers. Yet this same discretion also empowered caseworkers to mitigate the effects of poverty by providing individualized supports to deserving welfare recipients.

Since the second half of the twentieth century, welfare programs have increasingly targeted both people living in poverty and front-line workers as subjects of regulation. As welfare rights movements gained momentum in the 1960s, caseworkers’ discretion became a central concern of rights advocates and policy makers (Nadasen Reference Nadasen2005; Tani Reference Tani2016), and broader critiques of administrative discretion gained traction (Davis Reference Davis1969; Ontario Reference James Chalmers McRuer1968). Welfare rights advocates challenged caseworkers’ ability to deny benefits absent fair hearings or based on discriminatory notions of deservingness. Government officials, by contrast, were concerned with guaranteeing consistent decisions and preventing caseworkers from distributing public benefits to undeserving individuals, rather than ensuring that caseworkers provided aid to every eligible applicant (Handler and Hollingsworth Reference Handler and Jane Hollingsworth1971). Ultimately, between the 1960s and the 1990s, formal legal mechanisms were introduced to regulate front-line workers. Caseworkers’ discretion was thus “legalized,” as a large body of written rules was introduced to structure caseworker decisions, and “judicialized,” as front-line decisions became reviewable by tribunals and courts (Jowell Reference Jowell1975).

This legalization and judicialization failed to eliminate caseworker discretion, however, and instead created new space in which discretion flourished. Legalization’s lasting impact on North American welfare programs, including OW, is visible in their intricately-layered statutory provisions, regulations, and policies. As a regulatory strategy, legalization presumes that discretion and law exist in a zero-sum relationship. Yet, as socio-legal researchers have shown, discretionary space often increases in proportion to the number of rules created to constrain it (Hawkins Reference Hawkins and Hawkins1992). Statutory language almost always contains vagueness or imprecision, and when vague rules overlap, discretionary room multiplies (Sainsbury Reference Sainsbury and Hawkins1992; Endicott Reference Endicott2002; Brodkin Reference Brodkin1997; Prottas Reference Prottas1979). Moreover, reforms to social benefits programs often serve divergent policy goals, which become embedded in legal rules. Rather than resolve the conflicts produced by this approach, legislators make administrators responsible for ensuring benefits programs function despite their contrary aims (Baldwin and Hawkins Reference Baldwin and Hawkins1984, 574; Brodkin Reference Brodkin2008; Mashaw Reference Mashaw1983). As noted in the introduction, OW laws are marked by this legislative practice, and research participants explained how they regularly balance conflicting objectives, such as assisting vulnerable individuals, promoting self-sufficiency, and minimizing government spending, when they make benefits decisions. Most participants did not identify OW’s many rules as significantly restricting their discretion; rather, rules offered creative interpretive space.

Judicialization also ineffectively restricted front-line workers’ discretion. Not only are few caseworker decisions legally reviewable, but benefits recipients rarely challenge those decisions that can be reviewed. OW legislation limits the external administrative review of caseworker decisions to a small set of basic benefits decisions, making virtually all supplementary benefits decisions unreviewable. ^{Footnote 4} Courts are loath to review administrative decisions that they identify as addressing “policy” or “legislative” matters (Sossin Reference Sossin, Hertogh and Halliday2004b). For those decisions that can be reviewed, empirical researchers have demonstrated that judicialization ineffectively regulates caseworker discretion because government benefits recipients so infrequently appeal decisions that reduce or eliminate their benefits (Lens Reference Lens2005; Halliday Reference Halliday2004). Formal legal mechanisms thus doubtfully constrain caseworker discretion. If anything, the growth of legal rules and limits on external review ensure that discretion persists within OW and similar benefits programs.

From the early 2000s on, OW and similar programs have increasingly relied on managerial rather than legal devices to regulate front-line workers, particularly those that deskill benefits delivery tasks. American scholars link the deskilling trend in welfare agencies to two phenomena: shifts in program delivery models, from a social work model to a legal bureaucratic model (Diller Reference Diller2000); and changes in hiring and training practices that favour data entry clerks over professional social workers (Oberfield Reference Oberfield2014). My research suggests, however, that Canadian programs such as OW have pursued deskilling largely through new regulatory technologies rather than program delivery models or hiring practices. Although OW legislation introduced new principles to guide the delivery of last-resort assistance (e.g., promoting paid employment, conserving taxpayer-funded benefits), local offices’ approach to service delivery was mixed. Unlike some American welfare providers, caseworkers were not replaced with data entry clerks. In the offices I studied, front-line workers were highly trained. They possessed post-secondary diplomas or degrees in social work, social sciences, or public services administration, and their employers encouraged them to undertake graduate-level or continuing education. ^{Footnote 5} Further, many participants described taking a “social work” approach and striving to make client-centred decisions that would ameliorate hardship, an approach supported by management staff. Deskilling still exists in the OW program, but it is pursued primarily through ubiquitous regulatory technologies that caseworkers use in their daily interactions with clients.

While regulatory technologies may help caseworkers to manage legalization’s effects, specifically, the complex interplay between laws, regulations, and policies, they also promote deskilling by redirecting caseworkers’ attention away from clients and OW laws and towards check-boxes, flowcharts, and drop-down menus. Regulatory tools ranging from paper forms to computer programs have ballooned since OW was introduced in the late 1990s. For example, caseworkers must take every new OW applicant through provincial Application for Assistance and Rights and Responsibilities forms. Both documents prompt caseworkers to inform their clients about the range of benefits that they may be eligible for and their obligation to participate in employment preparation activities. Similarly, caseworkers are required to use a co-residency questionnaire to determine whether unmarried clients who live in shared housing are in a spouse-like relationship with their roommate. This form includes a series of questions that caseworkers must ask about the financial, social, and familial relations between clients and their roommates. Caseworkers must then decide, based on their clients’ answers, whether they and their roommate are in a dependent, “marriage-like” relationship, a finding which will ultimately reduce or terminate a client’s OW benefits. ^{Footnote 6} Although many participants found these tools helped them to navigate OW’s layered rules, some noted that form-filling distracts them from reading OW policies or spending time with their clients and that forms fail to capture either the nuances of OW laws or the realities of clients’ lives.

In addition to forms and questionnaires, computer software has become a popular tool to nudge caseworkers away from client-centred social work and towards data entry and caseload management tasks. For several decades, case management software has supplemented paper-based tools to guide the everyday decisions of those delivering social benefits. Rather than direct caseworkers to a particular outcome, however, early software regulated their decisions by organizing and revealing particular client data to caseworkers (Sossin Reference Sossin2004a; Herd and Mitchell Reference Herd and Mitchell2002). For instance, the case management software that preceded SAMS stored and displayed caseworker-entered data, enabling caseworkers to record notes about their clients, track OW payments, and perform basic benefits calculations (Herd, Mitchell, and Lightman Reference Herd, Mitchell and Lightman2005; “Rachel” Interview; “Bridget” Interview). This software organized client data and benefits payments but left substantive eligibility decisions to the workers themselves. Municipal social services departments have also developed their own digital tools to help caseworkers manage large caseloads and determine which benefits and services to grant clients facing multiple barriers to stable, paid employment. ^{Footnote 7} These programs prompt caseworkers to enter specific information about their clients’ education, training, and progress towards securing employment. They then translate this data into percentage scores that influence how caseworkers link individual clients to specific benefits. To help caseworkers manage their large caseloads, another program organizes client data so that workers can identify all clients in a particular office who are seeking a specific type of job or vocational training so that they can inform these clients of relevant upcoming job fairs or training opportunities (“Harriet” Interview; “Monica” Interview). These technologies subtly regulate front-line workers, however, because their functioning depends on caseworkers to conscientiously enter and maintain vast amounts of client data. Thus, they adjust caseworkers’ responsibilities from social workers who develop individualized, client-supportive solutions to inputters and managers of massive client databases. Nonetheless, caseworkers still maintain their role as gatekeepers of OW benefits, employment opportunities, training, and other community-based programs. As the next section demonstrates, the same cannot be said of the newest regulatory technologies, such as SAMS, which decentre the authority and expertise of those at the front-lines of OW delivery.

Regulating Generosity, Decentring Caseworkers

The new regulatory technologies introduced into social benefits programs more forcefully deskill front-line workers as they preoccupy caseworkers with data entry tasks and destabilize caseworkers’ decision-making authority. Although SAMS and its off-the-shelf relatives are marketed as effective regulatory tools based on their enhanced ability to audit caseworker decisions, their prescriptive design thwarts transparency goals and eclipses caseworker discretion. Presently, SAMS’s greatest regulatory impact stems not from its auditing capabilities but from how it displaces caseworkers as authoritative legal decision-makers.

The rationale for introducing SAMS—to decentre caseworkers as decision-makers—raises deep normative questions, which are gestured towards in the provincial Auditor General’s report that precipitated SAMS. In this report, the Auditor General expressed fear that municipal caseworkers too readily believe their clients’ claims of financial need, too broadly interpret OW laws in their clients’ favour, and too quickly distribute provincial funds to benefits recipients (AG Ontario 2009). A series of principled concerns underlie these fears. How broadly or narrowly should OW legislation be interpreted, keeping its conflicting policy goals in mind? Relatedly, whose legislative interpretation should be taken as authoritative: front-line workers (many of whom are trained as social workers) or provincial auditors (who are accountants by training)? Although answering these questions is beyond the scope of this article, they may offer insights into SAMS’s complex regulatory effects.

The Auditor General’s pre-SAMS report constructed municipal caseworkers’ discretion as a problem in need of a solution. In this report, the Auditor General criticized caseworkers for interpreting and applying OW eligibility policies and documentation requirements too flexibly, for too often relying on their clients to supply the information needed to establish individual eligibility, and for too readily trusting the veracity of client-provided information (AG Ontario 2009, 256–57). The Auditor General was not concerned that caseworkers might mistakenly fail to provide benefits to eligible individuals, but that they were interpreting OW laws too generously and distributing provincial funds too freely. Program managers were described as lax, allowing OW rules to be waived and condoning local practices that appeared to depart from provincial directives. The Auditor General was particularly alarmed by competing municipal and provincial interpretations of broadly-worded OW laws and policies, which caseworkers read more charitably than MCSS officials (AG Ontario 2009, 260–63, 265). In contrast with 1990s-era reforms, which aimed to eliminate an ostensible mass of fraudulent benefits claims (Mosher Reference Mosher, Gavigan and Chunn2010), the Auditor General’s recommendations addressed a new threat to OW’s integrity: the overly-sympathetic, too-generous front-line decision-maker.

The Auditor General recommended that the MCSS constrain caseworker discretion in order to govern the OW program, and the MCSS proposed SAMS as the most effective response. Simplifying and reducing overlapping provincial policies, while helpful, would not address the Auditor General’s underlying concern that municipal workers were innovatively interpreting provincial laws and policies and liberally distributing OW benefits. ^{Footnote 8} This problem instead required a solution that would rein in caseworkers while enabling the province to verify municipal compliance with provincial laws and policies (AG Ontario 2011, 388). The MCSS introduced SAMS as a managerial-technical answer to the dilemma of too-generous caseworkers. Already widely used by social benefits programs worldwide, SAMS promised more extensive remote auditing of caseworkers’ files (AG Ontario 2011, 386–87; AG Ontario 2015, 471, 474) and a “hands-off” approach to decision-making. Designed to generate its own decisions based on caseworker-entered data, SAMS would ostensibly make easy, routine benefits decisions and free up workers to meet with more complex clients (IBM Corporation 2012).

To say SAMS failed to function as promised is an understatement. After going “live” in November 2014, SAMS began releasing benefits payments to individuals who were no longer eligible for OW and eliminating or significantly reducing payments to eligible OW recipients. Rather than freeing up caseworkers to focus on their most vulnerable clients, SAMS demanded more of their attention by requiring workers to enter extensive client information into multiple data fields buried deep within the software and to “click” multiple times to perform basic tasks (“André” Interview). To halt the havoc SAMS was wreaking on front-line workers and benefits recipients, one public sector union unsuccessfully sought a court injunction (Brennan Reference Brennan2014). SAMS remained, and front-line workers adapted their decision-making practices accordingly.

While SAMS’s initial, massive benefit payment errors were widely reported, its regulatory rationale has more fundamentally transformed front-line decision-making. As noted above, earlier software assisted caseworkers with the procedural elements of decision-making, such as navigating complex rules, tracking benefit payments, and following up with clients. SAMS is qualitatively different. Like other off-the-shelf software, SAMS purports to be a “fully automated service delivery model” that produces its own legal decisions after caseworkers have input a wealth of client information into multiple data fields (IBM Corporation 2012). Though SAMS may eventually enable closer audits of caseworkers’ decisions, the threat of external review has had a less immediate regulatory effect on caseworker decision-making. Instead, SAMS’s performance as an automated decision-maker is the key to its regulatory impact. By acting as a decision-maker, SAMS transforms OW caseworkers from decision-makers who “mediate” between OW laws, office policies, social work norms, and clients’ needs (Prottas Reference Prottas1979, 149) into clerks who collect and input client data for SAMS’s use. In short, SAMS regulates primarily by displacing caseworkers from their vital role as legal decision-makers.

This regulatory rationale is evident in how SAMS functions. SAMS decentres front-line workers by firmly directing them towards a particular decision, which prevents workers from performing other essential decision-making tasks, such as assessing how many benefits a client has received, locating and reviewing notes that detail why a previous decision was made, and adjusting data inputs so that SAMS-generated decisions match a caseworker’s interpretation of client need and OW laws. As a software program, SAMS is designed to capture workers’ data inputs, analyze that data, and produce a benefits decision. As one participant described it, “[I]deally we’re supposed to input everything into that system, and the system will then produce the outcome: eligible or ineligible” (“Anita” Interview). Not only does SAMS generate its own decisions, but its design also makes it very time-consuming for front-line workers to insert themselves as decision-makers and influence SAMS’s outcomes. For instance, SAMS uses personal data to connect the files of present and previous OW recipients who, at one time, resided in the same household. It then uses this data to form families in ways that one participant described as being “almost like Ancestry.com” (“Martha” Interview). In doing so, SAMS decides that those individuals it has linked together depend on one another and, accordingly, reduces the total value of OW benefits provided to each “family” member. As a result, SAMS may make sole-support mothers dependent on previous household members, such as former intimate partners or their parents, even where caseworkers have reviewed relevant evidence and determined that these individuals do not live together (AG Ontario 2015). Front-line workers have great difficulty separating the people that SAMS joins because, as one participant noted, “SAMS doesn’t tell you, ‘Hey, this [data input] is what is causing your problem’” (“Stephanie” Interview).

SAMS also enforces a narrower interpretation of OW legislation, demanding time-intensive interventions by caseworkers who interpret legal rules more broadly. For instance, as noted above, benefit recipients are required to sign a Participation Agreement as a condition of receiving assistance (OReg 134/98, ss 18(1), 30). OW legislation and regulations are largely silent on the specific contents of these agreements and permit a broad range of activities to qualify as employment preparation activities, including vocational training, community service, and addictions treatment (OReg 134/98, ss 25–29). However, if a caseworker indicates in SAMS that a client has signed a Participation Agreement but fails to select a drop-down menu option specifying which employment activity their client has agreed to perform, SAMS will refuse to issue OW benefits. Similarly, though OW laws and policies empower caseworkers to provide clients with as many supplementary benefits as they are eligible for in one appointment, SAMS will not generate more than one supplementary benefit payment per client within a twenty four-hour period. One office manager explained how these conflicting interpretations of OW laws lead to particularly stark outcomes for benefits recipients. She described an individual who requested funds to purchase a shirt for an upcoming job-shadowing exercise. This client had received these funds on the previous day but, after buying a used shirt, he discovered that he was required to wear a crisp white shirt to the workplace:

While front-line workers may be able to “trick the system” so that SAMS provides benefits to legally eligible clients, this task requires that they enlist a team of colleagues, from supervisors to office managers, to work around or override SAMS. In other cases, workers may be able to wait twenty-four hours to issue benefits, but this means they must return regularly to a client’s SAMS file to issue one supplementary benefit at a time (“Rachel” Interview; “Cheryl” Interview; “Bridget” Interview). Given their large caseloads, caseworkers cannot tweak the system for every client and must selectively ration their efforts.

SAMS fractures, multiplies, and obscures client information and denies benefits if data are missing from its many evidence fields. These SAMS-generated decisions are very difficult for front-line workers to deviate from because the reasons for SAMS’ decisions are often hidden from caseworkers’ view. SAMS requires that caseworkers input the same client information into numerous boxes and screens, and clicking or not clicking a box can mean the difference between a client receiving or being denied benefits. Yet, when SAMS denies benefits to an individual whose caseworker has determined should be eligible for those benefits, workers must struggle to locate which data input is causing the problem. As one participant described: “[C]licking a checkbox means the difference between someone getting their cheque and someone not getting their cheque. So if you don’t click into seventeen different pages to actually find that checkbox, you won’t be able to figure out what is happening with this case” (“Julie” Interview). Further, SAMS buries data detailing the exact benefits payments that an individual has received while simultaneously preventing caseworkers from recording and saving notes to explain their reasons for particular benefit payments. The same participant explained:

Other participants noted that SAMS can make it unclear whether someone is even receiving assistance, as “it’s sometimes very difficult to see whether or not their file’s on suspend or if it’s closed” (“Jameela” Interview). Front-line workers cannot easily identify which types of benefits OW recipients have already received or the dates when these payments were made. These data are crucial for caseworkers to determine clients’ present eligibility for benefits. By obscuring basic information, SAMS bars those at the front lines from assessing whether individuals might be eligible for additional benefits and discourages caseworkers from issuing such benefits.

In addition to storing and displaying data in a way that fractures and obscures benefits decisions, SAMS also destabilizes caseworkers’ decision-making authority by hindering their ability to record reasons for a particular decision and multiplying the places where they might find one another’s notes. For instance, one participant observed:

Not only are notes recorded in multiple locations, but SAMS also has character limits built into its notes fields, suggesting through its design that detailed reasons for decisions are not required. In response to these limits, front-line workers may divide their notes across different fields. This practice produces fragmented lines of reasoning that can be difficult to revisit and for other workers to locate and review. Participants identified notes as a key obstacle to determining why previous benefits decisions had been made because the notes explaining a decision could be scattered throughout SAMS, recorded in a central notes database (the “Person Page”), or linked to one of many check-boxes, drop-down menus, or evidence fields associated with a particular benefit.

Finally, SAMS decision-making behaviour redirects front-line workers’ attention away from their clients and towards SAMS. Many participants described how they are now overwhelmed with administrative work and forced to use time they would prefer to spend meeting with clients or reviewing their caseloads to instead wrestle with SAMS (“Carys” Interview; “Dawn” Interview; AG Ontario 2015). For example, form printing is now so difficult that some caseworkers will suggest that their clients step out of a meeting so that the caseworker can ensure the proper forms print (“Sharon” Interview; “Rachel” Interview). This situation is antithetical to many caseworkers’ commitment to client-centred decision-making. As one participant put it, “Somebody’s crying and you have to say, ‘I have to ignore you for ten minutes and focus on this’ [makes typing sound with fingers on table]” (“Sharon” Interview). Additionally, because SAMS interprets client data unpredictably and reduces or cancels OW benefits unexpectedly, caseworkers often must review and try to correct SAMS-generated errors throughout their workday (“Nancy” Interview; “Stephanie” Interview). By demanding the undivided attention of time-pressed caseworkers, SAMS continually undermines them as skilled, client-focused decision-makers and repositions them as form-printers, data-inputters, and troubleshooters. As the next section suggests, however, by decentring and deskilling caseworkers, SAMS obscures rather than eliminates their discretion, prompting caseworkers to re-centre and re-skill themselves with complicated results.

Front-Line Discretion: Re-Centred and Black-Boxed

Front-line workers in the OW program respond to new regulatory technologies by asserting themselves as skilful social workers who are (or who become) adept at adjusting their data inputs to produce particular results. A dialectical relationship exists between SAMS’s performance as a regulatory tool and caseworkers’ response. As SAMS asserts itself as a decision-maker and transforms professional social workers into data-entry clerks, these workers adapt, re-skill, and re-position themselves. They learn to creatively interpret and enter client data so that SAMS generates outcomes that better match caseworkers’ perception of clients’ needs and their interpretation of OW laws and policies. Scholars have for some time now observed that regulatory technologies almost always risk being subverted by the subjects that they seek to regulate (Soss, Fording, and Schram Reference Soss, Fording and Schram2011a, 229). While these tools may black box caseworkers’ discretion, making it appear to outsiders that workers comply with rules and algorithms, in practice caseworkers find innovative ways to work with and around SAMS. As one participant put it, “It’s just the system—you have to know how to make it work to issue those funds” (“Dawn” Interview). Yet, even as front-line workers re-skill and re-centre themselves as professional decision-makers, they are not free to make decisions according to their own preferences (Baumgartner Reference Baumgartner and Hawkins1992; Mashaw Reference Mashaw1983, 213). Rather, caseworkers are repositioned in relation to SAMS; SAMS continues to function as a uniquely powerful institutional force governing their everyday decisions.

Front-line workers learn to adjust their decision-making practices so that their data inputs generate outcomes that more closely match caseworkers’ assessments of the benefits and services to which a particular client is eligible. Thus, even as SAMS directs caseworkers towards particular decisions and displaces them as decision-makers, workers find ways to redirect SAMS so that their clients receive the benefits they are entitled to according to flexibly-worded OW legislation. My data suggest three ways in which caseworkers creatively use SAMS to produce particular outcomes: first, by entering placeholder data when SAMS requires information that is impossible to provide; second, by adjusting dates forward and backward to moderate SAMS’s exacting interpretation of dates; and, third, by strategically categorizing clients’ needs so that SAMS will find these clients eligible for benefits.

First, caseworkers input placeholder data in fields where SAMS demands information that clients cannot provide. Though OW laws do not require this data to determine benefits eligibility, SAMS may require that very specific client data be entered. If these fields are left blank, or if a drop-down menu option is left unselected, SAMS will halt its benefits payments. To prevent SAMS from deeming clients ineligible for OW, front-line workers input placeholder information into these fields even if such information is inaccurate. For instance, SAMS requires that every adult OW recipient be enrolled in or have graduated from high school and that every dependent child be attending school, regardless of age. Similarly, SAMS requires every OW recipient to have an address, even those who are homeless or precariously housed. In response to these requirements, front-line workers have established a practice of marking all adult OW recipients as high school graduates, even if they have not yet received a diploma, and entering “fake school” into the school data field for children who are too young to be enrolled as students (AG Ontario 2015, 487). For clients who are homeless, caseworkers will input a fictitious address so that SAMS issues basic benefits (“Sharon” Interview; “Dawn” Interview). SAMS also requires that clients have a specific employment activity selected from the drop-down menu options in its Participation Agreement screen (which SAMS refers to as an “Outcome Plan”) as a condition of receiving benefits. As noted above, OW laws only require that individuals make a general commitment to engage in employment-preparation activities, but if specific details about a client’s activities are not input into SAMS by a set date, the software will find that individual ineligible for OW and stop issuing their monthly benefits. To counter this effect, workers may select any activity from SAMS’s drop-down menu options—life stabilization, finding employment, training, maintaining employment, restriction, deferral, and so on—to ensure benefits are issued, even if their selection misrepresents a client’s circumstances (“Bridget” Interview; “Stephanie” Interview). According to one participant,

Given their high caseload numbers, it is impossible for caseworkers to meet with all of their clients before the date on which SAMS will suspend payments. Front-line workers thus tweak the data in a client’s computerized file and make a reminder note to update this data at a future client meeting. To an auditor, this caseworker’s file may appear to be in compliance with SAMS’s requirements; the caseworker’s creative use of placeholder data would likely remain undetectable unless an auditor examines the caseworker’s note. Because these notes are scattered throughout SAMS, however, it is almost impossible for outsiders to comprehensively review this sort of decision.

Second, to temper the exacting way in which SAMS interprets event data, caseworkers will modify the dates they enter into SAMS so that their clients receive the maximum amount of benefits possible or so that benefits erroneously paid to a client are not clawed back from that client’s future OW payments. For instance, when a client has a new child, front-line workers may backdate when they add the infant as a dependent to that client’s file so that this client receives a larger monthly benefits payment. One participant described this process as follows:

Other front-line workers identified situations where they would adjust dates forward to protect a client’s benefits. When faced with a decision about whether to note an overpayment on a client’s file, participants described balancing provincial overpayment policies, SAMS’s data requirements, and an office norm to not create unnecessary hardship for their clients. According to one participant:

These adjustments to client data may lead a client to receive a small additional benefits payment. In this situation, OW’s legal framework does not offer caseworkers a single, correct answer; it grants them discretionary space. If we accept SAMS’s interpretation of OW law as authoritative, these data tweaks appear to be problematic. However, the participants who described making these adjustments interpreted OW laws differently. They noted that OW benefits are far lower than the actual costs of rent, food, and clothing, and that small additional payments could mean that clients are able to purchase groceries, pay utility bills, or remain housed. Though many participants described being guided by a general principle of preventing hardship, they also saw these data adjustment decisions as supporting clients’ progress towards self-sufficiency, one of the many goals underlying OW legislation.

Finally, front-line workers become adept at rearticulating their clients’ needs in language (or data) that SAMS recognizes so that SAMS will issue supplementary benefits. In such cases, workers negotiate with SAMS to ensure that clients receive the supports that their caseworkers judge to be harmonious with OW’s competing aims, such as providing clients with necessary assistance, promoting self-sufficiency, and prudently spending public funds. This rearticulation may require clients’ collaboration so caseworkers can provide them with much-needed benefits or services. When clients request funds to pay for an everyday item—a child’s clothing or school supplies, for example—that OW does not cover and that SAMS will not issue, caseworkers may look for alternate benefits for which these clients are eligible. After locating these benefits, caseworkers will input the data SAMS requires to grant these benefits on the understanding that their clients will use these funds to pay for the items for which they initially sought assistance. One participant stated:

In these situations, front-line workers demonstrate their proficiency as navigators of the numerous supplementary benefits available within the OW program and their skilful balancing of competing norms such as preventing client hardship and promoting self-sufficiency. Similarly, OW recipients in addiction treatment programs may require clothing funds that do not correspond with the drop-down menu options available in SAMS. In such cases, front-line workers may choose the next-best drop-down category so that their client will receive some funding. In the words of one participant who works with clients in addictions treatment:

Addictions treatment is one of the employment-preparation activities that legally entitles an OW recipient to supplementary benefits, such as a clothing allowance (OReg 134/98, s 26). However, because SAMS categorizes benefits more narrowly than OW legislation, its interface makes certain funds appear to be unavailable. Caseworkers who aim to provide clients with as many benefits as they may be entitled to under OW legislation must become adept at learning SAMS’s language, obtaining the evidence that SAMS requires from their clients, and cleverly inputting this data into the appropriate fields so that SAMS will generate a particular decision. In this sense, SAMS black boxes front-line workers’ discretion, as the reasons for a caseworker’s decision are not easily discernable to office managers, auditors, other caseworkers, or OW recipients.

Through their responses to SAMS, front-line workers re-skill and re-centre themselves as decision-makers, yet their discretion is distinctly shaped by and exercised in relation to SAMS. In the dialectical back-and-forth between caseworkers and SAMS, caseworkers cannot easily dismiss or evade SAMS’s centrality as a decision-maker. The lengths to which caseworkers will go in their creative responses to SAMS seem to depend on their commitment to social work norms (such as preventing hardship, or reaching client-centred decisions), but an equally important determinant is the limited time that workers have available for each client, especially as any deviation from a SAMS-imposed decision requires that caseworkers undertake more onerous data entry tasks, diverting their attention away from their other clients. Further, because SAMS threatens to cancel the benefits of individuals whose data are incomplete, it forces caseworkers to continually engage with its check-boxes and drop-down menus. By threatening to withhold benefit payments from vulnerable OW recipients if caseworkers overlook a single data entry, SAMS uses caseworkers’ commitment to preventing hardship against them. The regulatory pressures that this technology places on front-line workers, especially those who are dedicated to professional social work principles, leads workers away from their clients to the system itself. Even as caseworkers learn its language and adjust clients’ data so that they can more effectively “manipulate the system” (“Stephanie” Interview), “the system” continues to regulate them in response.