Long-Term Care Facility Workers’ Perceptions of the Impact of Subcontracting on their Conditions of Work and the Quality of Care: A Qualitative Study in British Columbia, Canada

Abstract Subcontracting long-term care (LTC), whereby facilities contracted with third party agencies to provide care to residents, became widespread in British Columbia after 2002. This qualitative study aimed to understand the impact of subcontracting from the perspective of care workers. We interviewed 11 care workers employed in subcontracted facilities to explore their perceptions of caring and working under these conditions. Our overarching finding was one of loss. Care workers lost wages, benefits, security, and voice. Their working conditions worsened, with workload and turnover increasing, resulting in a loss of experienced staff and a loss of time to provide care. These findings call into question the promises of quality and flexibility that legitimated policies permitting subcontracting, while adding to the mounting evidence that subcontracting LTC harms both workers and residents.


Introduction
This article examines the effects that the subcontracting of care has on the conditions for work and care in long-term care facilities (LTCFs) for older persons. The quality of care in Canadian facilities continues to be a concern (Armstrong & Day, 2017). In the public imagination, LTCFs have long been considered to be places of last resort and are associated with warehousing, waiting to die, and, during the COVID-19 pandemic, death itself (Ambrose, 2020;Estabrooks et al., 2020;Keefe, 2019;Vladeck, 2003). Cost has also been of concern, particularly in the context of an aging population wherein the need for long-term care (LTC) is expected to increase. These concerns have also come at a time of governmental fiscal restraint and greater reliance on market mechanisms to deliver seniors' care (Armstrong & Armstrong, 2020b).
One significant consequence of cost concerns has been a turn to the for-profit sector for innovation and investment. In the early 2000s, the province of British Columbia passed several "business-friendly" policies to prepare for increasing for-profit participation in LTCFs. Chief among these were policies permitting the subcontracting of LTCF staff. Subcontracting was a process whereby for-profit companies and non-profit organizations, which had been contracted by the regional public health authority to provide LTCF services, could decide to contract with third-party companies for their staffing needs. In British Columbia, this second layer of contracting was called "subcontracting". Facility management were able to end these subcontracts with short notice, allowing management to bust unions and reduce labour costs, a process termed "contract flipping" (Longhurst, Ponder, & McGregor, 2020). The governments' policy decisions to allow subcontracting were legitimated by claims that such arrangements would ensure financial sustainability, while supporting greater flexibility, responsiveness and quality (British Columbia Ministry of Skills Development, 2003).
Given the claims of improved service delivery, this study aimed to understand the effects of subcontracting from the perspectives of care workers themselves. We ask, how has subcontracting affected their working conditions and their capacity to care? Has it enabled more flexible, responsive caring? In what follows, we situate this study within the research on privatization in the LTCF sector. We then present a brief overview of privatization in the province of British Columbia, before turning to an examination of workers' experiences, drawing on interviews with 11 care workers from LTCFs who were subcontracted, some multiple times.

Privatization in the LTCF Sector
There has been a notable shift towards market-oriented approaches to health care financing and delivery over the past 40 years in Canada and other high-income countries. Popularized by the Thatcher and Reagan governments in the United Kingdom and the United States, this market ethos has been part of the neoliberal transformation of welfare states (Braedley & Luxton, 2010;Harvey, 2005). Although geographically variegated, this transformation has meant the increasing privatization of public services, such as the direct sale of public assets to private owners, public-private partnerships, contracting out, and increasing labour market "flexibility" by reducing the bargaining power of unions (Peck & Tickell, 2002).
"Privatization" refers to the transfer of public resources to forprofit providers (Mercille & Murphy, 2017). In neoliberal discourse, the turn to for-profit provision, in general and for the delivery of health care in particular, has been justified by the claim that market-based approaches are more responsive to public needs (Armstrong, 2010;Braedley & Luxton, 2010;Connell, 2010). In other words, for-profit providers are believed to adapt to changes in customer demand faster than governments, because they are motivated by the potential for increased profit (and equally by the possibility of financial losses). Such profit maximizing behavior is also thought to contribute to a more cost-efficient delivery of services. Neoliberal discourse also maintains that these gains will not be at the expense of quality. Indeed, quality is expected to improve through the mechanisms of provider competition and consumer choice, with poor providers being forced to improve or fail.
LTC has been particularly susceptible to privatization because the sector is excluded from the Canada Health Act (Jansen, 2011). This exclusion has contributed to the considerable variability in facility ownership across the country and within provinces (e.g., differing public, non-profit, and for-profit ownership mixes). Despite this variability, general trends point towards an increasing reliance on for-profit providers for the construction of new facilities as well as a concentration of for-profit ownership through large corporate chains (Longhurst, 2020, pp. 29-36;McGregor & Ronald, 2011).
In a discussion of privatization in the Canadian LTCF context, Armstrong, Armstrong, MacDonald, and Doupe (2020) observe that data availability on ownership is inconsistent across the country, making it difficult to finely track patterns in facility ownership. Presently, the province of Ontario has the highest proportion of for-profit provision (57%), with the majority (81%) of these homes being owned by corporate chains. Another 26 per cent of homes in Ontario are non-profit and only 16 per cent are publicly owned. By contrast, all LTCF in Newfoundland and Labrador are publicly owned (Government of New Foundland and Labrador, 2020). In the province of our study, ownership is almost evenly distributed. Currently, 38 per cent of homes in British Columbia are operated by health authorities, 34 per cent are owned by for-profit organizations and 28 per cent are owned by non-profit organizations (Office of Seniors Advocate, 2020). However, the difficulty in obtaining reliable data on ownership is concerning and suggests, as Armstrong et al. (2020, pp. 89) observe, that "there is little official interest in supporting the investigation of how different ownership patterns influence care." In addition to ownership, another important means of privatization is ideological, whereby market-based values, concepts, and practices are imported into the LTCF sector (Armstrong & Armstrong, 2020b), such as new public management (NPM) and auditing. NPM is a management style that emphasizes competition, efficiency, cost control, and quantitative outcomes (Rankin & Campbell, 2006). It is not particularly well suited to the hands-on, labourintensive LTCF environment wherein much of what is good carefor example, sitting and chatting with residents or taking the time to assist their independent eatingis difficult to measure. Related to NPM, the growth of auditing is another example of ideological privatization in the LTCF sector. As a practice imported from the financial sector, the turn to auditing to ensure quality has not fit the care context well (Banerjee, 2013). It has led to an intensification of regulation, standardization, and documentation while dramatically transforming the LTCF work environment, often in ways that make responsive caring more difficult (Banerjee, Armstrong, Daly, Armstrong, & Braedley, 2015).
The contracting out of labour is another significant means through which the sector has been privatized. "Contracting out" refers to the delivery of public services by for-profit agencies (Vrangbaek, Petersen, & Hjelmar, 2015). Typically, in the LTCF sector this has involved the contracting out of management or support services because these are perceived to be outside of care and, therefore, less important and less regulated (Armstrong & Day, 2017). For example, housekeeping, maintenance, food services, and laundry are commonly referred to as "ancillary" or "hospitality" services and are generally not regulated in the way that care work is. Although we do not agree that this work sits outside care in the LTCF context, the case of British Columbia is distinctive precisely because provincial regulations allowed for the privatization of direct care services. As noted, in British Columbia, the practice is referred to as "subcontracting" rather than "contracting out," because the facilities that these staff work in have already been contracted out from the public sector. This terminology thus captures the fact that subcontracting is an additional layer of privatization and profit extraction. In this article, we use the term "subcontracting" when referring to the process in British Columbia, and "contracting out" when referring to literature that employs this term.
Although changes in ownership, management culture, and the subcontracting/contracting out of staff are visible forms of privatization, the LTCF sector has and continues to be privatized through less obvious means. Austerity measures and insufficient public spending have resulted in an inadequate supply of LTC, creating lengthy waiting lists for beds and requiring citizens to resort to private means of support such as assisted living or other supportive housing arrangements. The subcontracting of staff is also indirectly caused by insufficient government funding. In a British Columbia study by , key informants reported that nonprofit employers wanted to maintain their in-house staff but were forced to subcontract because they could not afford to continue to employ the in-house staff given current levels of funding. Similarly, the hiring of private companions by families concerned with the quality of care that their relatives are receiving is another means by which LTCFs are being privatized (Daly, 2007;Daly & Armstrong, 2016). These various forms of privatization are referred to by Armstrong and Armstrong (2020b, p. 27) as "privatization by stealth" and share a common cause: a failure on the part of government to "respond to the need for public health care services and to supply appropriate care within care homes." The myriad forms that privatization takes within the LTCF sector has made assessing its effects challenging. Making comparisons among facilities at the level of ownership, for example, is made difficult by the layers of privatization. For example, nonprofit homes may subcontract some of their staff and thus be partially privatized. Moreover, non-profit facilities and even publicly owned facilities are not immune to the influences of market logic, either in their operations or their regulation. Comparing the quality of care between such facilities and for-profit facilities would therefore not provide a clear understanding of the effects of forprofit provision, because the boundaries between for-profits, nonprofits, and public facilities may be blurred. Nevertheless, the available research, which we will review briefly, has called into question the neoliberal rhetoric that for-profit provision leads to improved quality and efficiency.

Effects of Privatization
The weight of available evidence indicates that for-profit LTCFs deliver poorer quality care than either non-profit or public facilities (McGregor & Ronald, 2011;Poss, McGrail, McGregor, & Ronald, 2020;Ronald, McGregor, Harrington, Pollock, & Lexchin, 2016). The main reason for this difference is that for-profit facilities make their profits by keeping staffing levels and wages low and opting for staffing mixes with less-experienced staff (e.g., replacing registered nurses [RNs] with licensed practical nurses [LPNs], and LPNs with care aides). A large meta-analysis of studies comparing quality in non-and for-profit facilities found that non-profit facilities provided more and higher quality staffing than for-profit facilities (Comondore et al., 2009), leading the authors to conclude that residents would have received more care in non-profit homes. Similarly, a study by McGregor et al. (2005) found that the number of hours per resident-day provided by direct care staff was significantly higher in non-profit facilities than in for-profit facilities. And a more recent study by Hsu, Berta, Coyte, and Laporte (2016) found that for-profit homes, especially those owned by corporate chains, provided significantly fewer hours of care than non-profit or publicly owned homes. These lower staffing levels are significant, as staffing levels are strongly associated with quality in LTCFs (Schnelle, Jurgis, & Katz, 2013). Additionally, lower staffing levels and lower wages have been associated with higher turnover (Castle & Engberg, 2005), which fractures care workers' relationships with residents and their colleagues.
Research on contracting out staff has reached similar conclusions. In a systematic review of international studies published between 2000 and 2012, Vrangbaek et al. (2015) found that contracting out has predominantly negative consequences for workers. The studies that they reviewed reported that for-profit employers offered shorter-term contracts, which resulted in insecurity, relatively high turnover, and the loss of experienced staff. Not only did the workforce composition deteriorate, but working conditions also worsened, with studies reporting lower salaries, fewer benefits, and reduced job satisfaction. The research that Vrangbaek et al. (2015) reviewed also found that full-time employment tended to be replaced with part-time contracts, consistent with the rationale of workforce "flexibility". Work intensification was also reported in a number of studies, in most cases following reductions in staffing, resulting in "high pace and excessive workloads for the employees" (Vrangbaek et al., 2015, p. 15).
Whereas Vrangbaek et al. (2015) examined contracting out across the public sector, research focusing on the health care sector has also found deteriorating working conditions and negative changes to workforce composition (Armstrong & Armstrong, 2020a). In a study of the contracting out of health support workers, Stinson, Pollak, and Cohen (2005, p. 2) reported that newly privatized jobs were "substandard in all respects: low pay, meagre benefits, heavy workloads, poor training, and no job security." They also found that the combination of heavy workload, insecurity, inadequate compensation, and emotional stress not only had negative repercussions for workers but also significantly harmed their family life as well as their engagement within the community. Because health support work is highly gendered and racialized, the authors concluded that contracting out has turned back the clock on hard won gains in labour equity, with a privatized health support job being "virtually synonymous with poverty" (Stinson et al., 2005, p. 3).
Researchers have also raised (now prescient) cautions that contracting out staff could contribute to the spread of disease, noting that cleaning and laundry in the health sector is not at all like cleaning in the hospitality sector (Armstrong, Armstrong, & Scott-Dixon, 2008;Armstrong & Day, 2017). Special care must be taken when cleaning to prevent contagion from spreading. Also, the part-time contracts favoured by for-profit agencies intent on avoiding paying benefits force workers to work multiple jobs across different facilities in order to survive financially. This was foreseen to pose a risk for the transmission of disease (Longhurst, 2017), and indeed became a factor in the spread of COVID-19 in the LTCF sector (Estabrooks et al., 2020).

Privatization in British Columbia
Over the last three decades, the LTCF sector in British Columbia has been steadily privatized as a result of a number of provincial government policies. In the late 1990s, the government ended direct capital funding for the building of new care facilities. In its place, it implemented a competitive, request for proposal (RFP) process for all new publicly funded beds. Although the RFP process is, in theory, neutral vis a vis for-profit provision, the process is complex and favours well-capitalized organizations with the technical capacities and economies of scale to produce successful bids (Meagher & Szebehely, 2013). Therefore, between 2001 and 2016, the number of LTCF beds operated directly by regional health authorities and non-profits declined 11 per cent while beds in the for-profit sector increased 42 per cent (Longhurst, 2017). The growth of for-profit ownership between 1999 and 2020 is represented in Table 1.
The entry of private investment into the LTCF sector in British Columbia was further aided by two policy reforms in the early 266 Albert Banerjee et al. 2000s designed to reduce the power of unions and increase labour flexibility . Prior to this time, publicly funded LTCFs in the province were governed by a master collective agreement negotiated by an employer group (Health Employers Association of British Columbia) and a bargaining association representing unions. This agreement standardized working conditions and workers' rights across the province's LTCF sector. It generated a high degree of stability for the workforce where employment in the sector was seen as providing family-supporting wages.
The first policy reform, Bill 29, the Health and Social Services Delivery Improvement Act (hereafter, Bill 29), passed in 2002, allowed LTCF employers to lay off "non-clinical" staff, which included food service workers, housekeepers, and care aides. It also removed important union protections that ensured that collective agreements would follow the job. Previously, if part of a business was sold, transferred, or leased, the job would remain covered by the existing collective agreement. These "successor rights" had prevented employers from using subcontracting as strategy to dismantle negotiated contracts.
The second reform, Bill 94 the Health Sector Partnerships Agreement Act (hereafter, Bill 94) also passed in 2002, entrenched subcontracting in the sector and enabled employers to flip contracts. This meant that employers could terminate a contract with 60 days' notice, effectively firing the workforce, then re-tender the contract to a subcontractor, with no collective agreement following. The subcontractor could then hire workers, often the very same people who had been fired, with no union contract. Workers would need to organize and negotiate a new collective agreement, knowing that they could be fired again with 60 days' notice.
These bills resulted in waves of subcontracting, with more than 8,000 health care workers fired (British Columbia Ministry of Health, 2018), most of them women working in laundry, housekeeping, and food service occupations. For example, a press release from the Hospital Employees Union (2007) protested the mass firing of 450 care aides from three facilities in the Lower Mainland. That same month, 168 frontline care staff were fired on Vancouver Island, just weeks after signing a collective agreement. It was the third time in 3 years that staff at this facility had been fired (Hospital Employees Union, 2007).
Since these changes in the early 2000s, numerous studies have raised concerns about the effects of privatization for the people working and living in LTCFs (Longhurst, 2017; Office of Seniors Advocate, 2020; Stinson et al., 2005). Motivated by public complaints about the quality of care that residents were receiving, the British Columbia Ombudsperson undertook a province-wide review of the long-term facility and community care system. Among the significant problems identified was the harm done by the mass firing of staff. As the British Columbia Ombudsperson (2012, p. 368) observed: Such turnovers can disrupt the lives of seniors in residential care, especially those residents whose care needs are complex. Over time, long-term staff acquire specialized knowledge of these needs so the simultaneous replacement of many employees can make it difficult for the seniors because continuity of care is disrupted. This is particularly the case for residents with dementia. It can also be stressful to families since they often need to provide extra support to their relatives during such transitions.
Although the provincial government at the time did not heed concerns about the effects of subcontracting, an important response to the British Columbia Ombudspersons' inquiry was the creation of the Office of Seniors Advocate (OSA). The Office of Seniors Advocate (2020) recently completed an investigation raising serious concerns about privatization in the LTCF sector and its consequences for care. Echoing research on privatization in other jurisdictions, the OSA's investigation found that the forprofit operators provided less direct care and lower wages. Specifically, the OSA found that for-profit sector did not deliver the 207,000 hours of care that they were funded to deliver, even though they took more than $34,000,000 in profits. By contrast, the nonprofit sector provided 80,000 additional hours of direct care on top of the hours that they were funded to deliver. Further, the OSA found that for-profit operators paid less than non-profit operators, spending an average of 17 per cent less per worked hour, with wages paid to care aide staff being as much as 28 per cent below the union standard rate.
It could be argued, the Office of Seniors Advocate (2020) observed, that finding staff to do the same work for less and pocketing the savings is exactly what for-profit operators are expected to do. However, the OSA went on to question "whether the delivery of direct care hours in publicly funded long-term care homes is where we want operators to find efficiency" (Office of Seniors Advocate, 2020. p. 28). The OSA further speculated that subcontracting might result in higher turnover, the loss of experienced staff, and recruitment problems, while disrupting the continuity of care. We will present findings from a qualitative study wherein care workers speak to these concerns about the consequences of subcontracting.

Methods
We designed this study to learn about the effects of subcontracting in LTCFs from the perspective of care staff whose voices have been typically absent in the making of LTCF policies. Ethics approval was obtained from the University of British Columbia (H17-02162).
We worked in coordination with a major LTCF sector union to recruit participants. We sent two recruitment e-mails to union members requesting volunteers to take part in interviews. Only two responses were received. This low response rate may be attributable, as our interviews later discovered, to the nondisclosure agreements that employers required staff to sign as part of their contract, which prohibited them from discussing their work.
A second recruitment strategy was attempted. We contacted facility union stewards and asked them to identify staff willing to participate in the study. Eleven participants were identified and their contact information was obtained. Dr. Ponder contacted interested participants and conducted interviews over the phone. She completed all interviews but one in 2017. She obtained voluntary consent at the start of the interview after a brief description of the study. During the interview, she asked participants to describe the process of subcontracting and its consequences as they experienced them. Interviews were audio recorded and transcribed verbatim with identifying information removed. We analyzed the data to identify patterns pertaining to the practice of subcontracting and its effects on work and care (Kvale, 2008). However, we note that given the small sample size, we did not analyze data by facility ownership or jurisdiction, as anonymization protocols stipulated that participants would not be linked to place of employment.
Our analysis took place in several inter-related phases. First, Dr. Banerjee conducted close readings of initial transcripts and then used open coding to develop a coding scheme. The research team discussed and refined the coding scheme. Then Dr. Banerjee coded all interviews using this scheme. Through this process, he identified a number of preliminary themes pertaining to the experience of subcontracting. Second, he compiled summary case notes for each participant pulling together relevant data on these themes. Third, he re-analyzed these case summaries, comparing and contrasting across participants to identify patterns of commonality and difference. The developing analysis was then shared with the other authors for discussion and further refinement. Finally, Dr. Banerjee conducted a negative case analysis of the original transcripts to ensure that the analysis offered a faithful representation of participants' experiences and perspectives (Shenton, 2004).

Sample Description
Participants included one RN, three LPNs, and seven health care aides (HCAs). All participants were female except for one HCA who was male. Although we gathered data on gender, we did not gather data on country of birth or race. Participants had work experience in the LTCF sector ranging between 5 and 28 years. They worked in facilities across three southwestern British Columbia Health Authority regions: Vancouver Island Health, Vancouver Coastal Health, and Fraser Health. Eight participants were working in for-profit facilities and three were working in non-profit facilities. These publicly funded facilities had all subcontracted with for-profit companies to provide their direct care services. Nine interview participants were working at these facilities before and after contracts were flipped. Several facilities had been subcontracted multiple times, with one facility having had contracts flipped four times. Several participants had worked at LTCFs that had not been subcontracted and some continued to work at such facilities, providing them with additional points on which to compare the conditions across subcontracted and non-subcontracted sites.

Findings
In what follows we describe the consequences of subcontracting from the perspective of care workers. None described subcontracting positively. Indeed, our overarching finding was one of loss. Subcontracting was experienced as a series of losses to workers' roles as care providers (e.g., loss of wages and benefits), worsening of their conditions of work, and losses to their ability to care for residents. We follow this thematic organization in the presentation of our data, describing the impact of subcontracting on workers, the conditions of work, and the conditions for care.

Reported Impacts of Subcontracting on Care Workers
Subcontracting resulted in losses to workers' compensation and job security. Wages decreased after contract flipping and were significantly lower than those for similar positions at non-subcontracted facilities. Scheduled pay raises were reduced. One worker (RN1) characterised the new pay raise structure as "criminally low" (RN1), and another (HCA2) remarked that 3 years after subcontracting, she was still earning $4 per hour less than her starting wage at a non-subcontracted facility. Workers further reported losing shift "differentials", which were described as the higher wages earned for working "off hours", such as overnight or on the weekend. HCA5, for example, reported her weekend wages decreasing by a third: from $24 to $16 per hour.
Participants lost benefits. Reductions in health and dental coverage were reported. Pension contributions were reduced or lost entirely. And workers received significantly fewer sick days. For example, HCA5 noted that her sick days were cut from 18 days to 5. Participants routinely described the effects of subcontracting through the language of loss: "We lost everything. We lost the shift differential, and there's no pension plan. There's a very minimal health care plan" (RN1).
Workers also lost voice. When signing new contracts, interviewees reported that they were required to sign nondisclosure agreements. These agreements were not solely concerned with protecting residents' privacy. Rather they were described as "mostly to do with not talking to the media about what goes on in the facility" (RN1). As a consequence of these nondisclosure agreements, workers lost agency and their capacity to campaign for better working conditions and improved care for residents.
The thing is a lot of us right, most of us unfortunately are bound by the confidentiality agreements that we signed. We can't go to social media. We want to campaign for a union campaign we've got which is a really good one. [It's a] 6-week campaign. We want it to go viral. We can't. We can't. These people have armies of lawyers. They can afford armies of lawyers with the public dollars that they've got behind them (LPN3).
Subcontracting was used as a tactic to dismantle existing unions, and the threat of mass firings discouraged further organizing. For example, rather than bargain with the existing union, LPN1 reported that the employer fired all staff and re-tendered the contract. In the process, staff lost the gains that they had achieved in their previous contract. Another participant mentioned that staff at her facility considered unionizing, but felt it was pointless; if they succeeded, the company would "just fire all of us" (HCA1). Staff also lost their most vocal members, with activist staff and prounion workers fired and not re-hired.
In sum, from the perspective of staff, subcontracting resulted in significant losses in pay, raises, benefits, and agency, while increasing their financial vulnerability. Indeed, vulnerability appeared to be a condition for remaining. When asked why they stayed, participants described themselves as being unsure of finding employment elsewhere given their age or familial responsibilities that made it difficult to start at a new home with no seniority. In the case of a 63-year-old participant (HCA5), he explained that he believed that he was not only too old to find work elsewhere, but also that his language skills were poor and posed a significant employment barrier. He felt that he had no choice but to take the new contract. Some participants also added that they remained out of concern for residents.

Reported Impacts of Subcontracting on Conditions of Work
According to all participants, working conditions deteriorated substantially after subcontracting. One consequence of the lower remuneration previously mentioned was the loss of experienced workers. "We lost a lot of really good employees who decided not to stay on with the new staffing company," observed LPN1. "We've never regained that level of expertise in our employees. Now we work with a really low level of expertise." Another consequence attributed to the poor remuneration was the difficulty in hiring capable staff. According to participants, new hires were very inexperienced, and their low level of competence was concerning. "There's a lot of people working there that are just, just totally inappropriate to be working. It's created a kind of dangerous situation. I have several people that I just refuse to work with because they're so under-qualified to be there" (LPN1). The hiring of inexperienced staff increased workloads, as HCA6 explained.
It's really hard…I have these new care aides. It's their first time to work. It is so hard for me because I do most of the work. I am so tired because I still need to [explain it] and then do it [myself]. Oh my God I am so tired to say it. "Don't do this one! You know what to do, that's why you're here, right?" It's overwhelming! The loss of expertise was not a temporary setback, as noted by LPN1. Indeed, several participants observed that they lost the time and energy that they had invested in training new hires because the new hires could not be retained, and the facility lost the experience that these new recruits had garnered. As RN1 explained: "They work for a couple months. They get a reference. And they go work for [names a nonsubcontracted facility]." HCA2 expressed it succinctly: "We are just a revolving door." Subcontracting led to an increase in workloads. This increase was attributed to several factors beyond the additional work created by inexperienced hires. Participants reported that staffing companies did not hire as many workers, resulting in an increase in the number of residents that workers were responsible for. For example, LPN1 observed that the ratio of residents to LPNs doubled in her facility. Another facility increased the acuity level of residents but did not increase staffing levels commensurately, which contributed to greater workloads.
Short-staffing also increased following subcontracting, further increasing workloads. This was not an exception but routine. "We are always short staffed. Always. Every single day" (HCA1). Short staffing was a vicious circle not only contributing to the poor working conditions but also exacerbated by them, because workers who could avoid coming in did so. For participants who had employment elsewhere, working at the subcontracted facility became a last resort "Especially now that our contract [end] is getting close, the anxiety is very high," observed LPN3. "This is why I think we're having so many overtimes. Because some people go for a second job now, they spend more time at their second job because they feel that they have more security there rather than working here." Workers reported that calling in sick to avoid picking up shifts was common. Even casual workers were described as avoiding these facilities, which added to the loss of staff and overwork, as HCA3 explained.
They cannot staff all the lines. It's because although we have lots of casuals on the list. There're no casuals that like to work here. Because these casuals have other jobs in other facilities that pay more. So, of course the prioritythe last priorityis us.
Many of those we interviewed felt that the subcontracting companies did not respect care work. HCA2 illustrated this through the HCA appreciation day at her subcontracted facility. Because it was only one day, if you were not scheduled to work on that day, you would miss it. By contrast, the non-subcontracted facility where she worked held an appreciation week so that all staff could be sure to attend. Another contrast provided by HCA2 was how her managers responded to incidents of violence from residents. At the subcontracted facility, she was blamed for causing the incident and received a written reprimand. At the other facility, the manager personally called her, apologized for the incident, and expressed concern for her well-being. She recounted receiving the phone call while driving and the difference was so stark that it brought her to tears: "I pulled over on the side of the road and I just cried. I was like I can't even believe how different it was and like how crappily we were treated at the other place." Similar accounts of distress caused by the losses experienced through subcontracting were reported by all participants. RN1 summarized these sentiments.
[The flip] was a really, really big morality hit for everybody. Everybody was really upset….It was really negative. It was heartbreaking. We lost a lot of good staff and then a lot of benefits… It was just loss after loss and then the working conditions got worse because we hadall the suddenwe had no staff. It's been a real loss…

Reported Impacts of Subcontracting on Care
Losses were also reported with regard to workers' ability to provide care. The intensified workloads meant lost time with residents. Under conditions of heightened work pace, participants said that they did not have the time to spend with residents and therefore were not able to understand what residents might be needing in the moment. As HCA1 explained: "There's no one-on-one time. There's no time to figure out how [residents are] doing or anything. You can't even like learn about them except for reading their files, if you get time." The increased turnover also affected workers' ability to develop familiarity with both the residents and their colleagues, further hampering teamwork and care. The loss of time meant that workers' focus was on completing tasks. Responsive care was sacrificed, as illustrated by HCA1.
[At the subcontracted site I work at] we have no wake policy. Because you'd be so behind; it's pretty much people get up whether they want to get up or not! You fight with them. Whatever. Just get them up! [At the non-subcontracted site where I work] they tell you to take your time. However long it takes you to get a resident done, that's fine. (HCA1) Workers also reported losses in the areas of communication and coordination. For example, the addition of a subcontracting company created mixed allegiances, wherein subcontracted care workers were not employed by and therefore not responsible to the facility but rather to the subcontractor. Power struggles impeding the coordination of care resulted. For example, HCA5 described a situation wherein a housekeeper was asked if he could clean up a carpet because a resident had vomited on it. He responded to the nurse, saying that "I cannot go back because I finished already there." In another account, a housekeeper was vacuuming the carpets while the residents were eating and a nurse asked if he could do it later so that residents could eat peacefully. The housekeeper responded: "You are not my boss! My boss is [anonymized]; and she's the only person who can ask me to stop vacuuming now." Given the loss of staff and time, participants resorted to workarounds as a means of coping with unmanageable conditions. These workarounds could be risky and could put workers and residents in harm's way. The creation of a medication assistant position was an example. This involved providing an HCA with some training to administer medications, freeing up time for LPNs. However, giving HCAs responsibility to administer medications that "they did not understand" was described as "very dangerous" and resulted in "medication errors" (LPN1). Moreover, this workaround reduced the time that nurses spent with residents. The act of administering medication was an important moment of connection between LPNs and residents, offering them an opportunity to conduct a "little assessment" and to remain informed about the medications that residents were taking. This workaround thus fostered disconnection. "I didn't even know they were on antibiotics because I didn't look at their medication for four days" (LPN1) Other harmful workarounds included hiring casual staff instead of permanent staff, pulling workers from lighter care areas to work in heavier areas, using students to fill gaps, or requiring a nurse to supervise the entire facility rather than only one unit. Participants reported feeling distressed because their working conditions prevented them from providing the care that they knew they could and should be providing. You cut corners. No one gets a bath. People stay in bed. People stay wet because you don't have time to change them. If you get to them by the end of your eight hour shift that's considered good! I just, I can't feel good about myself working at a place like that. HCA2 Participants reported losing the ability to keep residents safe. "At the end of the day we face a lot of risk of violence and aggression, because we just can't handle it. The needs of our residents are so high. We just can't keep everybody safe right now" (HCA1). Participants' accounts attributed the cause of violence not to the residents' conditions but to their conditions of work. As RN1 remarked We have one locked unit with 20 residents that have very advanced dementia but they're also aggressive. So they all lump them together in that unit. Well that unit isn't staffed with any more people. They don't get anymore. They get hardly any recreation time. So they're all walking around like the walking dead. They're bumping into each other and they're hitting each other. They're fighting each other, pulling each other out of bed. It's a free-for-all basically.
Workers were also put in harm's way. RN1 went on to add: "[Residents] are aggressive to staff. They're hitting, kicking, punching, biting, spitting. Everything that you can think of that you would do to somebody, they do to us."

Discussion and Conclusion
The subcontracting of direct care staff has been a distinct and deleterious means of privatizing the LTCF sector. The policy choices that enabled subcontracting were justified by promises of financial sustainability, greater responsiveness, and improved care. The authors of this study sought to assess these claims by learning about the consequences of subcontracting from care workers themselves. From the workers' vantage point, the overarching experience of subcontracting was one of loss. Care workers lost wages, benefits, job security and voice. Subcontracted care workers were not empowered to provide better care through more flexible working conditions. Rather, their working conditions worsened, which harmed their ability to adequately care for residents. Workers reported increases in violence and unsafe workarounds to cope with challenging conditions. Staff who were able to leave, did.
Although these findings of loss are at considerable odds with the rationales that legitimated subcontracting, they are congruent with studies that have attended to the perspective of employees, finding that contracting out staff worsens working conditions, disempowers workers, and reduces the quality of services provided (Vrangbaek et al., 2015).
Research on contracting out in the LTCF sector has also shown that the profits made tend to be taken from workers, through a number of tactics that keep labour costs down (Armstrong & Armstrong, 2020a;Stinson et al., 2005). Our findings add to this body of evidence. In no case did staff report improved compensation. Rather, subcontracting resulted in lower wages and the elimination or reduction of benefits, pensions, and sick leave. Shift differentials were also lost and pay raises were so low that staff could work for years without approaching their starting salary at facilities that employed staff directly.
Returning to the question posed by the Office of the Senior's Advocate (2020, p. 28) as to "whether the delivery of direct care hours in publicly funded long-term care homes is where we want operators to find efficiency…," our results suggest the answer is no. Not only were workers harmed, but also subcontracting did not result in greater efficiency. Rather, less care was provided less well. As both the British Columbia Ombudsperson (2012) and the Office of the Senior's Advocate (2020) feared and our findings confirmed, subcontracting damaged relationships between staff and residents. Mass firings, turnover, and overwork made it difficult for workers to become familiar with residents and have the time to respond adequately to residents' needs. To these findings, our study adds rich stories from care workers about the further disruptions among staff and between staff and management. For example, tensions were introduced between facility management and staff members who worked directly for the subcontracting agency and who were, therefore, not responsible to facility management. These mixed allegiances resulted in conflicts that disrupted resident care.
Our study also suggests that these disruptions were not temporary. Workers reported that, as a consequence of the low remuneration and poor working conditions, their facilities became "revolving doors" with new hires entering and leaving as soon as they could find better paying work elsewhere. The experience that was lost through the mass firings was not regained over time. Moreover, gaps in continuity of care were not adequately filled by casual staff. Not only were casual workers unfamiliar with the facilities' residents, staff, and work routines, they also they avoided working at subcontracted facilities whenever possible. In this way, the practice of subcontracting further contributed to a workforce of poorly paid workers needing to work at several facilities to support their families and with minimal sick benefits. Indeed, combined with frail older seniors susceptibility to COVID-19, these working conditions have been identified by public health officials as playing a major role in the spread of the virus in LTCFs (Longhurst & Strauss, 2020;McGregor, 2020).
One limitation of this study is its small sample size. However, as Morse (2000) has observed, small sample sizes can be acceptable in studies, such as this one, where the topic is focused and easily understood by participants. Another limitation is the possible selection bias of our recruitment strategy. Nonetheless, the attributions that staff made are logically coherent and serious enough to warrant concern. They are not atypical either. Our findings are commensurate with other research that has documented significant negative consequences for workers (Armstrong & Armstrong, 2020a;Stinson et al., 2005;Vrangbaek et al., 2015) Given the preponderance of negative consequences, one may wonder why the privatization of LTCF sector continues. One answer to that question is evidentiary. As Ronald et al. (2016) observe, the gold standard for research evidence is the doubleblind randomly controlled trial, which they note would be difficult, let alone unethical, to conduct in the LTCF context. Researchers must therefore resort to observational studies, and these studies of privatization are further complicated by the blurry boundaries between facilities that are run on a private for-profit basis and those that are not. In our study, two of the non-profit homes had privatized their workforce. Given these evidentiary challenges, Ronald and colleagues suggest that the precautionary principle be followed. Adopting a precautionary approach would shift the burden of proof and call for "preventative action, even when there is uncertainty but credible evidence of potentially significant impacts" (p. 8).
Another motive for the continued privatization of the sector worth considering is the political and economic rationale proposed by Mercille and Murphy (2017). They argue that given the dearth of empirical evidence for efficiency and effectiveness, privatization is better understood as a strategy to reduce the power of labour. With this in mind, the evidence of harm revealed by research such as ours makes more sense. Subcontracting suppressed worker compensation, eliminated collective agreements, and dissuaded workers from further union organization. These losses were not then a policy failure, but from Mercille and Murphy's standpoint, they were precisely the point of subcontracting.
This transfer of resources and power and the concomitant harms to both workers and residents have not gone unopposed in British Columbia. The work of advocacy groups, unions, activists, and researchers have engendered change. In 2017, the provincial government promised $500,000,000 in new funding, directed primarily at increasing staffing levels. Although, this investment was much needed, it did not address the legacy of privatization. However, after their election in 2017, the New Democratic Party (NDP) took on subcontracting directly. The government passed new legislationthe Health Sector Statutes Repeal Actwhich repealed the laws (Bills 29 and 94) that had granted operators the unrestricted ability to subcontract direct care staff and avoid union successor rights. The repeal of Bills 29 and 94 does not turn back the clock by removing subcontracting from the sector entirely. Many direct care staff continue as a subcontracted workforce with inferior collective agreements and fewer workplace rights than public sector care staff have.
Change can come from surprising places. At the time of writing, the COVID-19 pandemic is exposing "long-standing, wide-spread and pervasive deficiencies in the sector" according to a recent Royal Society of Canada report, provoked by the fact that more than 80 per cent of deaths in Canada were associated with the LTCF sector (Estabrooks et al., 2020). Responding to the pandemic, the British Columbia government has set about standardizing LTCF wages, bringing them up to the level of the wages in the master collective agreements in public health authority facilities. The government has also instituted a "one job, one facility policy" eliminating the practice of working across sites to cobble together a survivable income.
More recently, the provincial government committed to continuing the "'levelled up wages' even after the pandemic ends, and [restoring] provincial standards for wages, benefits, and working conditions" (British Columbia New Democratic Party, 2020, p. 20). It remains to be seen how government will achieve these policy objectives. Government could facilitate structural changes in the industry through a ban on subcontracting, and through requiring that for-profit operators re-join the existing provincial master collective agreement. Alternatively, the government could provide further subsidy to for-profit operators, similar to the approach taken in response to COVID-19 (Longhurst & Strauss, 2020). By removing the incentive to cut labour costs via subcontracting, the first approach would necessarily reduce the industry's profits. The second approach would require additional government funding even as many of these operators are already being funded at a level that assumes that they are paying the provincial unionized wage standard when they are not (Office of Seniors Advocate, 2020, p. 28). This would seem to nullify claims of the superior cost effectiveness of for-profit providers in the first place.
We conclude by observing that our study is the first to our knowledge to examine the working conditions of subcontracted British Columbia LTCF care staff from their perspective. The themes of lossreduction in wages, worsening working conditions, and disruptions in care deliveryarrive at a time when the extraction of maximum efficiency from the care-providing relationship is being shown to have fatal consequences.