Corps de l’article

Accommodating disability in the workplace can be messy and complicated. When an employee becomes disabled, the law requires employers to consider alternative arrangements, but meaningful, productive work that the employee can do isn’t always easy to find. Every reasonable effort must be made, short of undue hardship,[1] to accommodate. This means that the employer must do more than simply investigate whether an existing position might be suitable. Efforts must be made to determine whether other positions in the company might be appropriate, or whether a job description might be adapted to fit the employee’s capabilities. The employer must look at all other reasonable alternatives and the efforts must be both genuine and thorough (Lynk, 1998). Furthermore, the accommodation process must be individualized in order to meet the particular circumstances of each case. Therefore broad policies, such as collective agreement clauses that allow for termination of a disabled employee after a specified period of absence (i.e., two years), cannot be rigidly applied without consideration of all other factors.[2]

Unions also have an obligation to cooperate with accommodation, although the employer normally directs the process. They bear responsibility equal to that of the employer if discrimination arises as the result of a mutually negotiated provision of the collective agreement. If the union is not initially a party to the discrimination, the primary obligation remains with the employer. While accommodations that substantially interfere with the integrity of a collective agreement would likely exceed undue hardship parameters, it is also clear that the agreement may not be used as a shield against human rights.[3]

Sometimes when an employee returns to work, competence (or luck) reigns and the disabled individual progresses from modified to full duties or is permanently accommodated with relative ease. In other situations, however, the return-to-work (RTW) process is protracted and rife with conflict. The employee is shuffled from department to department or there are multiple attempts to accommodate that fail (Butler, Johnson, and Baldwin, 1995) ultimately resulting in dismissal or arbitration.

So what is the difference between these two scenarios? We know that modified work programs and length of service with the employer (Dasinger et al., 2001) can have an effect, as can physical and psychological demands of the job (Krause et al., 2001a, 2001b). But sometimes human beings are the ones who botch it up. So, the question is: how might we, unions and managers, be contributing to problematic accommodations and what can be done to increase the likelihood of success?

Literature Review

Studies examining problems between disabled employees, unions and managers are quite rare. Most investigate only the employee-employer relationship and can be separated into two broad categories: research on managerial attitude, and research on organizational culture.

Managerial Attitude

Field studies suggest that one of the biggest barriers to implementing disability legislation is managerial inflexibility (Cunningham and James, 1998). Although managers indicate they wish to cooperate, they perceive problems regarding financial and organizational limitations, a lack of knowledge about applicable legislation, and the inability to recognize signs of illness (Larsson and Gard, 2003; Cunningham and James, 1998; Lee and Newman, 1995). Jackson, Furnham and Willen (2000) surveyed managers from 200 companies and found that willingness to comply with legislation is associated with positive attitudes and greater knowledge of the law. When negative managerial attitudes do exist, they are entrenched and difficult to shift (Habeck et al., 1998).

Experimental work has found that stereotyping (Colella and Varma, 1999), onset control, past performance, and magnitude of an accommodation request (Florey and Harrison, 2000) also affect attitude. Furthermore, supervisors report higher quality relationships with disabled subordinates when the subordinate displays ingratiating behaviours (Colella and Varma, 2001).

Disabled employee assessments of managerial attitude have also been polled. Attitudes valued by employees include: responsiveness, validation, fairness, shared decision-making and managerial empathy (Shaw et al., 2003). Maintaining contact with absent employees and informing coworkers of possible changes in task assignments is also important (Nordqvist, Holmqvist, and Alexanderson, 2003). When negative attitudes prevail, employees believe managers may try to get rid of them (Slack, 2000) or may use resistance strategies to discourage them from requesting accommodation (Harlan and Robert, 1998). Low levels of supervisor support are also associated with reduced RTW rates (Krause et al., 2001b), increased lost time (Feuerstein et al., 2001) and increased use of sick time (Bourbonnais et al., 2005). However, some reviews have suggested that evidence of the impact of managerial support is not conclusive (Geertje, 2001).

Only one study was found that sought perceptions from multiple participants. Ponak and Morris (1998) interviewed disabled workers, managers, co-workers and union representatives. Three themes were identified: the importance of communication and information; ambivalent attitudes towards accommodation; and conflict over scarce resources.

Organizational Culture

Hunt and Habeck (1993) were the first to provide a framework for evaluating contextual workplace factors affecting RTW. Managers were asked to rate their organizations’ performance on a series of policies and practices. Those that correlated highly with workers’ compensation claim outcomes were distilled into three categories: 1) safety management and prevention, 2) a comprehensive disability management system, and 3) organizational climate. Follow-up work by Habeck et al. (1998) added detail and particularly recognized the impact of an adversarial climate and lack of trust.

Amick III et al. (2000) tested the Hunt and Habeck (1993) questionnaire on 158 disabled employees. The two dimensions most relevant to this research tested “people-oriented culture” and “labour-management climate.” In a factor analysis, the labour-management climate variables did not load significantly. However, people-oriented culture positively predicted RTW. Subsequent research confirmed the culture effect (Cullen et al., 2005), and validated the Amick III et al. (2000) scales for use with both managers and employees (Ossman et al., 2005).

Because research on accommodation is quite new, there are many gaps in the literature. First, there are few theoretical models. Stone and Colella (1996) developed the most comprehensive framework based on an integration of work from sociology and psychology. But, because the model is based on theories from other disciplines rather than field observation, there may be some variables that remain undiscovered. Further, the explanatory mechanisms may not reflect common conditions in the workplace. For example, Stone and Colella (1996) assume that the disabled employee is capable of performing assigned tasks. Research from Butler et al. (1995) indicates that this is often not the case.

A second concern is that assessing RTW from a tripartite perspective is very rare. While unionization is sometimes assessed as a single dichotomous variable in disability studies (Butler, Johnson and Baldwin, 1995; Habeck et al., 1991), and there is substantial practitioner literature on union-management cooperation (see Jodoin and Harder, 2004), the complex relationships that exist remain unexplored in empirical work. This seems particularly important considering that 29.7% of Canadians are unionized (Akeampong, 2006).

Methods

This study addresses some of the gaps identified above. Field research was conducted using arbitration cases, in-depth interviews and grounded theory methods of analysis to identify tripartite relationship factors that negatively impact the RTW.

Data collection proceeded in two phases. Phase I involved examining data from arbitration awards in three Canadian jurisdictions. In a unionized workplace, grievance arbitration is the legally mandated system for resolving disputes that arise out of a collective agreement, such as discipline and dismissal.[4] Recent Supreme Court decisions have made it clear that arbitrators have the jurisdiction (when the essential character of the dispute arises from the collective agreement) to consider human rights legislation in their decisions.[5] This means that human rights issues have come increasingly under the scrutiny of labour arbitrators, who now adjudicate a large proportion of accommodation cases.

For this study, seventy-two cases (see the Arbitration Case List at the end of the article) were analyzed, not for legal content, but as a source of historical and observational data. Arbitration cases were chosen because the narratives provide unobtrusive information on the interaction between disabled workers, their unions and employers. The historical records are accurate, vetted by neutral arbitrators, and include data that may not otherwise be easily accessed (e.g., physician testimony, disciplinary records, etc.). Arbitration cases have been recognized as a uniquely reliable data source for investigating organizational policies and practices (Zerbe, 2005).

The labour arbitration case sample was selected from Quicklaw databases for Alberta Grievance Arbitration Awards (A.G.A.A.), Ontario Labour Arbitration Awards (O.L.A.A.) and British Columbia Collective Agreement Arbitration Awards (B.C.C.A.A.). The key words “disability” and “accommodation” were used to search the databases. Only cases with disability accommodation as the principal cause of the grievance were included.

The Alberta sample, encompassing all 31 cases reported between January 1996 and June 2002, was used to develop preliminary categories. Ten of a possible 44 cases from Ontario, reported between January 2001 and June 2002, were then randomly selected and coded for inter-rater reliability and identification of additional categories. Another nine Ontario cases were purposefully selected for rich descriptive content. Two researchers independently analyzed each of the nineteen Ontario cases in order to compare findings. Finally, a fourth sample of 22 British Columbia cases (of a possible 189 cases reported between 1993 and 2002) were randomly selected to broaden representation and improve generalizability.[6]

A detailed coding guide of 93 variables was then developed and tested. All 72 cases were re-coded by three auxiliary raters. Just over 1,500 single-spaced pages of documentation were analyzed for 76 grievors. Table 1 provides descriptive statistics for age, sector, and decision by province.

Table 1

Descriptive Statistics for Age, Sector and Decisions by Province

Descriptive Statistics

Male

Female

Public

Private

Upheld

Denied

Alberta

15

20%

16

21%

12

16%

20

26%

13

17%

20

26%

Ontario

10

13%

6

8%

7

9%

12

16%

8

11%

10

13%

British Columbia

17

22%

7

9%

12

16%

13

17%

10

13%

14

18%

Total

42

55%

29

38%

31

41%

45

59%

31

41%

44

58%

Gender information was unavailable for five grievors

Upheld = upheld in whole or in part

One decision was reserved

-> Voir la liste des tableaux

Phase II involved in-depth interviews (conducted between January 2002 and September 2003) with managers, union representatives, occupational health workers and disabled employees who had previously experienced or were currently undertaking RTW. Participants were identified and approached through occupational health departments, unions, or senior management. Secondary data was collected in the form of collective agreements, policies, procedures and RTW documentation.

A semi-structured interview protocol was developed using the categories distilled from the arbitration cases. Participants were asked to identify relationship issues that significantly contributed to the outcomes of accommodations. They were asked to recount incidents that differentiated between successful and unsuccessful accommodations. Recalled fact situations and perceptions were recorded and transcribed. Interviews were done in person, and lasted from 60 to 180 minutes. A total of 23 interviews were conducted with 22 participants from both private and public enterprises.

A limitation of the research is reliance on a large sample of arbitration cases. This may introduce bias in that arbitration cases are a product of controversy and therefore may under represent good practices. In order to balance this inequity, efforts were made to include participants who had experience with successful accommodations. It should be noted that despite this effort, participants were less likely to describe positive incidents unless prodded.

Bias may also be introduced because arbitrators are reporting events through a legal lens which may screen out important contextual elements. Furthermore, although less likely to engage in impression management, arbitrators are “as vulnerable as other experienced professionals to the operation of implicit theories of human behaviour” (Zerbe, 2005: 21).[7]

Another limitation is that data was collected only in unionized settings because the investigation targeted the tripartite relationship. Restricting data collection to unionized populations means that the findings may not be generalizable to managerial staff and non-union populations.

Lastly, it should be noted that the incidence of disability in this study may be exaggerated. Participating managers estimated that the number of accommodated employees on their units at any given point in time ranged from 20% to 40%. Occupations in both the interview and arbitration samples were often physically demanding or repetitive (see Table 2 for arbitration case job titles).

Table 2

Job Descriptions

Job Descriptions

Number

%

Mechanical/Operations

19

25

Health Care

15

20

Warehouse/Transportation

9

12

Office/Clerical

6

8

Education

5

6

Janitorial

4

5

Cashier

4

5

Other

8

11

Missing

6

8

Total

76

100

-> Voir la liste des tableaux

Results

Four tripartite relationship factors that contribute to accommodation difficulty emerged from the analysis. The first category, managerial attitude, suggests that bias may stem from questions of employee credibility and then spill over to affect future accommodations. The second category, employee involvement, indicates that excluding the disabled employee from accommodation planning occurs frequently and has a negative affect on communication patterns. The third category, ineffective investigation, highlights the difficulty managers have balancing illness legitimacy with confidentiality requirements. The final category, union-management climate, looks at union roles in accommodation and suggests that substantial union-management animosity taints RTW efforts.

A detailed exploration of each category follows, with fact situations and interview quotes used to illustrate the findings. Where detail in the cases allowed for coding, simple descriptive statistics provide additional support to the qualitative scrutiny.

Managerial Attitude

The first factor that emerged from the analysis was how frequently accommodated employees sensed that they were unwelcome. Interviews with managers indicated that autonomy with respect to accommodation arrangements can present an opportunity to exercise preferential treatment. So some workers are able to reap the benefits of exceptional accommodations, while others are forced to accept inflexible and unfavourable positions:

Let’s not kid ourselves. You could have the best employee…and you might try to bend some of the rules. Yet I’ve got six people over here that I forced out of work because they don’t fit the system right now.

Manager

Evidence from the arbitration cases supported managerial testimony. In 66% (31/47) of the cases where information was available, there was some indication of managerial reluctance to accommodate, and the existence or perception of bias contributed to accommodation difficulty. The cases were coded as follows:

  1. Where there was a direct statement indicating that the employer was particularly helpful in the accommodation (e.g., “the employer went the extra mile” Goodyear Canada, 1998, para. 77), the case was coded as: bias not a factor. This occurred in 16 cases.

  2. If reluctance or bias was perceived by any participant in the case, but other factors had a greater effect on accommodation outcomes, the case was coded as: bias a minor factor (17 cases).

  3. When managerial bias or reluctance contributed substantially to problems with accommodation, the case was coded as: bias a major factor (14 cases).

  4. Policy grievances or cases with insufficient testimony to determine the existence of bias were coded as missing (27 cases).

In some cases managerial bias appeared early on, influencing the quality of accommodation offered (e.g., Finning, 1995; United Nurses of Alberta, 1997). For example, in Finning (1995) there were indications that managers used a minor disability to fabricate an opportunity to terminate the employee. The grievor had 21 years seniority and a relatively minor, but chronic, wrist injury. The employer offered to accommodate the grievor in the lowest paying and least secure position in the company. The offer included the demand that he “perform the full tour of duties without exception” (Finning, 1995: 4). Union officials believed the employer was setting the grievor up for failure, and that other more suitable accommodation was available but the offer was deliberately harsh because of managerial bias. The arbitrator agreed with the union, and the grievance was upheld (Finning, 1995).

Reluctance and bias were heightened where managers questioned the employee’s credibility or the legitimacy of the illness (e.g., Goodyear Canada Inc., 1998; Providence Healthcare, 2001; 35, Westmin Resources Ltd., 1998). In several cases managers believed that employees were attempting to “scam the system.” Scamming was identified when a disability claim was raised late in the day in order to avoid transfer, discipline, or reclassification (e.g., Canada Safeway Ltd., 2000; Imperial Oil, 2001; Transalta Utilities Corp., 1998). Managers also assumed scamming existed when the employee was off work and receiving disability benefits, but was “moonlighting” at another company (e.g., Communication, Energy and Paperworkers Union, 2000; Woodbine Entertainment Group, 2001). Scamming was suspected in ten cases. In all but one of these instances, the employee’s claim of illness was subsequently substantiated.

Sometimes credibility was questioned even if the behaviours were illness-based. Although medically explained, the behaviours impacted managerial assessments because they existed prior to diagnosis or because they were particularly disruptive. Disruptive behaviours that were determined to be a direct result of a disease process included absenteeism (Mill and Timber Products, 1993), falling asleep on the job (Slater Steels, 2001), impairment from drugs or alcohol (Health Employers’ Association of British Columbia, 2000; Nestle Canada Inc., 2001), and erratic or inappropriate conduct (London City, 2001; Shuswap Lake General Hospital, 2002).

Whether assessments of employee trustworthiness were accurate or not, suspicion tended to spill over, damaging all RTW systems. As one manager explained: “the legitimately injured employees get a bad name because people paint them all with the same brush.” As a result, accommodations were sometimes designed as a form of vigilante justice. Managers would make disparaging comments regarding productivity or punish the employee by assigning the least desirable shifts and duties: “if you are on modified duties you should work the night shift or…you can do it on the weekends or you can do it at 4:00 in the morning.” [Occupational Health Worker].

Further, there is evidence that managerial expressions of enthusiasm for the accommodation plan and welcome for the returning employee were observed and mimicked by coworkers, thereby potentially amplifying the effects:

A lot of it depends on the manager: the manager’s body language, the manager’s attitude. We’ve got some that are really good and their staff will buy into it more than the manager who’s really reluctant.

Union Representative

Even managers who readily acceded to the ethical principles of accommodation and acknowledged employee credibility, sometimes found it difficult to accept the obligation. This was most likely to occur when the accommodation was for a worker from another area, or when the requirements arose frequently or unexpectedly.

These managers were influenced by the practical challenges associated with accommodation. The time, energy and “creativity” necessary for developing an accommodation plan were often cited as barriers. Managers also indicated that building cohesive, stable teams became more difficult, especially if accommodation requests were generated externally. There was discomfort exhibited with the very personal nature of a disability scenario. As one occupational health worker observed: “they don’t like the nitty-gritty stuff.” It was also suggested that if managers who accommodate are not rewarded for their efforts, or become responsible for the added costs, there is little motivation to divert time and energy from other activities.

Finally, it should be noted that employee perceptions of managerial reluctance may develop even when bias does not appear to exist. For example, in the Ontario Nurses Association (2001) case, there was no permanently available modified work for the grievor because two other employees were already being accommodated. Yet the employee was convinced the organization had not done all it possibly could to find an alternative: “I could not understand why a modified position could not be created for me when they had already created it for two other staff members” (OntarioNursesAssociation, 2001: 7).

Employee Involvement

The second theme that emerged was the frequent exclusion of the disabled employee from accommodation planning. The exclusion is not necessarily intentional. However, the evidence suggests that if it is not explicitly incorporated into the employer’s roles and responsibilities, it is likely to be left unattended.

There were a total of 42 arbitration cases in which there was some attempt to return the employee to work, and where there was sufficient information to determine the degree of involvement. The level of disabled employee involvement in accommodation planning for the arbitration cases was assessed by coders as very involved (4 cases), somewhat involved (13 cases), not at all involved (25 cases) or missing (30 cases). This means that in about 60% of unsuccessful accommodations where a return to work was attempted, the central figure in the process was excluded from employer meetings regarding work options or even from scheduling appointments for third party medical assessments (e.g., Communications, EnergyandPaperworkersUnion, 2000).

Simple cross tabulations (see Table 3) indicate that men and women are involved in RTW planning relatively equally (42% of women had some level of involvement compared to 40% of men). Public sector organizations are more likely to involve their employees than private sector companies, and Alberta lags behind the other provinces with an involvement rate of only 18%, compared to Ontario at 42% and British Columbia at 70%. The cross tabulations also show that while communication errors are actually more likely to occur with greater involvement, the fact that communication exists at all seems to be associated with lower rates of re-injuring subsequent to a return.

If we assume that greater involvement will generate successful accommodations, there should be some explanation for high involvement/low success stories. Looking at the cases with employees who were very involved, two common threads can be found that may explain the paradox. First, these were cases where the employee was attempting to exercise substantial control over the accommodation process. For example, in CanadaSafeway (1998), the employee had requested multiple transfers and was characterized by managers and peers as “a know-it-all.” In the TorontoCity (2001) case, the employee had a lifelong condition and was pushing the employer to allow her to do more than her physician had indicated was appropriate. In the high involvement cases there were also many attempts to accommodate the grievor (more than six in three of the cases), and none of the grievances were upheld, indicating that although the accommodation was unsuccessful, the employer’s efforts met the undue hardship threshold.

Employees who were somewhat involved experienced between one and six attempts to accommodate and were most likely to have success at arbitration (eight were upheld, five were denied). These employees were contacted by the employer, but this did not preclude communication problems. Discussions were frequently conducted over the telephone and occasionally a meeting including the disabled employee would be arranged. However, face-to-face meetings often occurred after the relationship had already become hostile. For example, in the Woodland Windows (1997) case, the majority of RTW discussion occurred via telephone, so offers of accommodation were not documented. The employee did not have a clear understanding that accommodations would be made and subsequently did not attend work. The employer assumed he was “deliberately milking the system” and suspended him for two weeks (WoodlandWindows, 1997, para. 56; see also OntarioNurses’Association, 2001; VSAHighwayMaintenance, 2002).

Table 3

Employee Involvement in Accommodation Planning

Involvement

Very

Somewhat

Not at All

Total

Male

1

4%

10

36%

17

61%

28

Female

3

21%

3

21%

8

57%

14

Total

 

 

 

 

 

 

42

Public Sector

2

18%

4

36%

5

45%

11

Private Sector

2

7%

9

30%

19

63%

30

Missing

 

 

 

 

 

 

1

Total

 

 

 

 

 

 

42

Alberta

1

6%

2

12%

14

82%

17

British Columbia

1

8%

8

62%

4

31%

13

Ontario

2

17%

3

25%

7

58%

12

Total

 

 

 

 

 

 

42

Frequent/Occasional Communication Errors

2

13%

6

40%

7

47%

15

Rare/No Communication Errors

2

7%

7

26%

18

67%

27

Total

 

 

 

 

 

 

42

Reinjured

0

0%

1

14%

6

86%

7

Not Reinjured

3

9%

11

34%

18

56%

32

Missing

 

 

 

 

 

 

3

Total

 

 

 

 

 

 

42

-> Voir la liste des tableaux

Finally, employees who were not at all involved experienced between zero and three accommodation attempts. Arbitrators upheld these grievances about fifty-five percent of the time.

So it appears that the quality of the involvement is as important as the degree. Employees who attempt to manipulate the accommodation process by demanding more than can be offered, or by refusing to accept expert advice may alienate individuals they need as allies. When employees do participate, it is usually at the employer’s discretion, so if things go awry because of ineffective efforts, the employer will be held responsible. Completely excluding employees is linked to fewer accommodation attempts and may be interpreted as the employer failing to exert a genuine effort and therefore falling short of legal requirements.

Evidence from the interviews further supported the arbitration case data. Where the employee was excluded, misunderstandings regarding the employee’s abilities, willingness to work, or the employer’s desire to have the employee back in the work force were cited as problems. Unless the employer invites the employee to attend planning discussions, the employee is unlikely to be aware of the options considered, production concerns, or scheduling limitations that impacted the final accommodation offer. This may inadvertently raise questions in the minds of even successfully accommodated employees regarding the level of effort or the quality of the accommodation. As this quote demonstrates, accommodated employees become highly attuned to the presence of other disabled workers and carefully compare the nature of different accommodations: “Some people, they can find light duties for and others are just supposed to go back to work….How does that work?”

Ineffective Investigation

The third theme was confusion on the part of both managers and employees regarding the balance between sufficient investigation of disability and employee harassment. In 64% of the cases where the grievances were upheld, arbitrators stated that the employers’ failure to properly investigate contributed to the conclusion that the duty to accommodate was not met. Another 14% of these cases were policy grievances, which means that only 22% of the companies in the sample had completed thorough investigations.

Arbitrators found that employers failed to gather sufficient data regarding the extent of the employee’s illness (MainlandSawmills, 2002; NiagaraStructuralSteel, 2001), the employee’s functional capacity (VSAHighwayMaintenance, 2002), or possible task adjustments (ShuswapLakeGeneralHospital, 2002; SaultAreaHospitals, 1999) prior to dismissing the disabled employee for either culpable or non-culpable reasons.

In other cases, supervisors who were suspicious of diagnosis validity sometimes over-investigated, crossing the confidentiality line with persistent and invasive requests for additional medical information (e.g., ProvidenceHealthcare, 2001; UnitedSteelworkersofAmerica, 2001). This pattern was repeated in the interview data as well. Employees and their physicians viewed the trend as “harassment,” managers saw it as necessary, and union representatives were concerned primarily with the invasion of privacy and additional stress that resulted from constant pressure to justify an illness.

Not all requests for information were driven by suspicion. Many managers were genuinely concerned about the employee’s welfare, or were seeking information to plan for the worker’s return or develop strategies to deal with longer-term absences. However, unless a manager is particularly skilled at communication or has a long term and positive relationship with the employee, there may be a perception that the employer is pressuring the employee to return too early: “When you go back to work on a modified, they can’t wait to get you back to work, they don’t give a shit what you do as long as you are back to work so it looks good on them” [Accommodated Employee].

Union-Management Climate

Collective agreements mean union acquiescence is often required before modified duties can be identified for disabled workers. While overt cooperation is mandated by law, this research suggests that accommodation outcomes may be substantially affected by the quality and degree of union involvement. The importance of this factor emerged through interviews as opposed to case analysis, and was considered by this occupational health nurse as one of the top two contributors to accommodation outcomes:

First of all I would say the earlier you get on a case, the better the outcome is going to be. Then I would say getting all the stakeholders on board, if there is a union try and get them to buy in.

One union representative compared two incidents to illustrate the importance of a coordinated effort. In the first situation the union representative, occupational health and management met with co-workers prior to the actual RTW date. The returning employee’s disability, although not named, was legitimized by the process. Work restrictions and the anticipated length of the accommodation were revealed along with a request that coworkers respect those limitations. It was also indicated that the accommodation was made with the union’s consent, and contact information was provided with an invitation to call should staff have any concerns.

In the second incident, another accommodation was required on the same unit. The manager attempted to independently repeat the exercise, but the outcome was described as a disaster. Disgruntled staff were “creating chaos.” The manager ultimately called in occupational health and the union to repair the damage.

When asked what she thought made the difference between the two scenarios, the union representative suggested that sometimes employees trust the union more than management, and that she personally used the discussion as an opportunity to educate staff: “I try to explain that the human rights legislation says that we have to accommodate…if it was you next then you would expect the union to assist you too.”

The story illustrates that unions may have a particularly influential role to play by nudging collective beliefs and norms about accommodation. Clearly establishing support for the accommodated employee, informing peers about the law, and imparting some empathy may help diffuse the animosity that results from disruptions to the status quo. The union was also a critical resource when conflict escalated. Managers reported that they called the union when the behaviour of employees with psychological disabilities became disruptive.

Other union roles recounted in the interviews that contributed to successful accommodations included: a) assisting disabled employees with workers’ compensation claims or private insurers, b) offering emotional support and guidance through administrative processes, c) maintaining contact with employees while they were off work for recovery, and d) pushing management to look harder when an appropriate solution was not readily apparent.

On the negative side, union-management animosity sometimes played a very destructive role by distorting communication and interfering with identification of accommodation options. In these organizations, unions were described by managers as political institutions with little concern for the welfare of the employees under their protection. Union officials were accused of obstructing accommodation attempts by refusing to cross-group employees or by advising employees not to release medical information. Mistrust of union motives was consistently expressed by managers who believed business rationale could be used to explain the more cautious approach to RTW that the union advocated:

They are running a business, their business is union dues. The more union dues they get the more money they make. They get more union dues by hiring more people…So if somebody’s off work and they can extend that absence, then we have to bring in more casual employees to fill in…. What you’ll hear is that they want to make sure that the person stays off until they are completely recovered, but it boils down to the dirty aspect, union dues.

While I saw no evidence that managers employed sabotage tactics because of union-management relations, mistrust spilled over to taint accommodation related interactions. First, as one manager revealed, employees who relied on union communication channels were assumed to be malingering: “rather than bringing in the [medical] form and give it to the supervisor, they will drop it off at the union office, it’s a hint they are trying to work around the system.” Second, it was suspected that employee physicians were being manipulated by unions. Third, the managerial animosity towards the union was picked up and echoed by other occupational health workers. Interviews with these participants were peppered with comments such as “unions keep themselves in business by creating crisis” and unions are “what schoolyard bullies grow up to be.”

Highly inflexible union-management relationships contributed to other communication obstacles between nurses and returning employees as well. It was believed by occupational health practitioners that filtering medical information through the union disrupted meetings and delayed accommodation planning: “they won’t come and see me unless they bring a union rep. And now I have been told that if they bring a union rep, I must have someone from management in the room.”

In some organizations the negative attitudes were so entrenched that developing strategies to overcome communication problems was unlikely. Even though it was recognized that union involvement could greatly enhance accommodation success, managerial and occupational health response more often meant excluding the union:

I have seen more and more what the union wants is to make management life miserable. And they will sometimes sacrifice the individual to do that. So that’s why I have decided to leave them out.

One of the most unfortunate outcomes of union exclusion was the perception by the accommodated employees that they had been abandoned at a difficult time, and that the union inactivity was deliberate. One employee explained the confusion she felt as the wheels fell off her accommodation, and the union failed to contact her: “fourteen years of working here and nobody came to my rescue.” Although employees were sometimes ambivalent about direct involvement with the union, they were clear they wanted to see proactive efforts in health and safety: “they should be there all the time, not just when something goes wrong.” In some cases they also expressed the belief that the union should represent the employee outside of the employer-employee relationship, where the union has no jurisdiction (e.g., in workers’ compensation or Canada Pension Plan disputes).

Clearly not all organizations that participated in the study experienced extremely strained union-management relations, but it should be emphasized that when they existed, they were pervasive. The construct repeatedly identified as critical for accommodation success, and most damaged by union-management animosity, was trust:

In a non-union environment you can speak to an employee and you can plan and listen and you believe the employee and they believe you. Here you can’t build up trust with employees.

Occupational Health Nurse

Finally, the existence and pervasiveness of union-management animosity was generally attributed to history. As one worker observed, “You know what they say, you get the union you deserve.” Some managers conceded the point as well, and explained that although there was a concerted effort being made to change the culture of the organization they were in, it was a difficult process because adversarial beliefs and responses were so ingrained. This made cooperation on disability accommodation a learning process with hurdles that were higher than might be seen in a less antagonistic setting:

The union is slowly changing, but it’s hard, just the same as it is hard for us to get some of the senior supervisors to change…They have been moulded under a very different philosophy.

Manager

Conclusion

This research was undertaken to find out how managers and unions might be contributing to difficult accommodations, and highlights four principal problems: managerial bias, exclusion of the disabled employee from planning, accommodation investigation errors, and strained union-management relations. While some very pragmatic conclusions can be drawn for managers and unions, the research also raises many questions and pinpoints potentially fertile areas for further investigation.

The first contributor to difficult accommodations, managerial bias, has been acknowledged in previous research (e.g., Bourbonnais et al., 2005; Jackson, Furnham and Willen, 2000). The unique contribution here comes from detailed content analysis which revealed that bias could stem from the additional burden of accommodation, or could originate with distrust that springs up quickly and spreads from one disability situation to the next. Bias that erupts from distrust is particularly interesting because distrust is a thread that wound its way through all four categories. It was a concern at both the individual level and at the organizational level, and begs the question: what is it about disability that so easily shifts our thinking?

Trust is well investigated in psychology and management literature (e.g., Mayer and Gavin, 2005; McAllister, 1995; Pugh, Skarlicki, and Passell, 2003), but has been linked to RTW outcomes in only one other study (Amick III et al., 2000). Trust affects perceptions of procedural fairness, decision commitment and attachment to work groups (Korsgaard, Schweiger, and Sapienza, 1995). It has also been shown that managers are more inclined to meet work-related needs for individuals they trust (McAllister, 1995).

This research shows that trust may be a powerful explanatory construct for disability accommodation. For example, previous studies have found that breaches of trust may be triggered by changing the rules after the fact, broken promises, wrong or unfair accusations, or disclosure of secrets (Bies and Tripp, 1996). Actions observed in the cases that may have been perceived as breaches of trust included: a) over-investigation of illness legitimacy, b) punishment of the accommodated employee through less desirable work assignments, c) lack of emotional support, d) scamming, and e) disruptive, illness-based behaviours.

Trust-breaching directly impacted accommodation by decreasing the likelihood that disabled employees would be offered positions appropriate to their skills and abilities. Furthermore, managerial distrust was noticed by coworkers who mimicked unsupportive behaviours, making the workplace psychologically uncomfortable for the returning employee. Breaches of trust also escalated conflict, which decreased union-management cooperation, impacted commitment to the RTW plan and, at its worst, lead to disciplinary action and expensive arbitration.

It is well beyond the scope of this paper to fully explore the relationship between trust and disability accommodation. However, the research results clearly indicate that trust has explanatory potential. A new workplace disability model that incorporates this important construct should be developed.

The second contributor to difficult accommodations that emerged was employee involvement. In cases that went to arbitration, employers often failed to engage the returning employee in accommodation planning. Involvement in and of itself was not sufficient to improve accommodation outcomes (or chances at arbitration); timing and quality were also important. In cases where employees were involved, employers were sometimes guilty of making contact only after important decisions had already been made, so the employee was unaware of the extent of the accommodation search.

When this occurred, returning employees relied on comparing their outcomes to those of other accommodated staff to determine whether or not their experiences and workloads were fair. This process of comparing outcomes suggests that another fruitful area of investigation may be equity theory. Colella, Paetzold, and Belliveau (2004) have begun some work in this area, testing co-worker assessments of procedural justice in disability accommodation. However, to my knowledge, there is no research that has assessed disabled employee perceptions of accommodation fairness, or the impact these assessments may have on RTW outcomes.

Another interesting finding in the employee involvement category was the significant variation between provinces, with Alberta having very low involvement and British Columbia very high. Human rights or employment law distinctions may account for some dissimilarity in business climate and employment practices. Another plausible explanation may be awareness due to educational initiatives (Harcourt, Lam and Harcourt, 2005). For example, British Columbia is home to the National Institute of Disability Management and Research, and the University of Northern British Columbia, which has the only graduate degree in disability management offered in Canada. In either case this research points toward macro environmental differences that warrant additional scrutiny.

The third contributor, poor management investigation is a new finding. It seems that a delicate balance must be found between inadequate investigation of accommodation options and overzealous pursuit of medical detail, particularly when a worker is feeling vulnerable. Under-investigation may be the result of inexperience or workload pressures. Over-investigation is more challenging. It occurred because legitimacy of the disability/work restrictions was questioned. This highlights again the importance of understanding how trust and disability interact in the workplace.

The final category, labour-managementclimate, indicated that unions may be a particularly important force in the evolution of collective beliefs about disability and fairness. Union presence may also have a damaging effect when the climate between the parties is so hostile that it spills over to accommodation. So labour relations makes a difference, but the impact is unlikely to show up if tested as a dichotomous variable (e.g. Butler, Johnson and Baldwin, 1995), or by asking only about cooperation (e.g. Amick III et al., 2000). Large scale studies that capture the degree and quality of union involvement may be more revealing.

Disability accommodation in the workplace is a real source of frustration for those in the trenches and research is just beginning. This paper reports on some of the factors that can be influenced by unions and employers, but also hints at the importance of issues such as perceptions of disability legitimacy, co-worker attitudes and the relationship between the employer and employee’s physician. More investigation is clearly needed and I would like to close with an invitation to labour and management scholars. A solid research foundation is needed before we can offer effective recommendations to those in the field.