PROJECT FINAL REPORT Workplace Incivility and Other Work Factors: Effects on Psychological Distress and Health

Principal Investigator Harry S. Shannon, PhD.

McMaster University

Co-Investigators Ted Haines, PhD. Lilia Cortina, PhD.

McMaster University University of Michigan

Research Team Lauren Griffith, PhD. Candidate Lacey Langlois, MSc. Candidate Vasudha Gupta, BSc. Candidate Kazi Othir Moitri, BSc. Candidate

McMaster University McMaster University McMaster University McMaster University

Prepared for the Canadian Union of Postal Workers July 3, 2007

Table of Contents SUMMARY................................................................................................................................................. IV INTRODUCTION ........................................................................................................................................ 1 BACKGROUND ............................................................................................................................................ 1 INCIVILITY .................................................................................................................................................. 1 GENERAL APPROACH ................................................................................................................................. 2 STUDY AIMS ............................................................................................................................................... 3 METHODS AND RESULTS....................................................................................................................... 4 STUDY POPULATION ................................................................................................................................... 4 SAMPLING: WHO WAS IN THE STUDY .......................................................................................................... 4 QUESTIONNAIRE ......................................................................................................................................... 5 Item Selection: What questions were asked .......................................................................................... 5 Pilot Test ............................................................................................................................................... 5 MAILING ..................................................................................................................................................... 6 Response Rate ....................................................................................................................................... 7 Demographic Characteristics ............................................................................................................... 7 DATA ENTRY .............................................................................................................................................. 7 DESCRIPTIVE ANALYSIS ............................................................................................................................. 8 Incivility and Sexual Harassment.......................................................................................................... 8 The Union............................................................................................................................................ 11 Work Environment .............................................................................................................................. 11 Health and Pain .................................................................................................................................. 12 Injuries and Safety .............................................................................................................................. 12 Discrimination .................................................................................................................................... 13 RELATIONSHIPS BETWEEN ‘EXPOSURES’ AND ‘OUTCOMES’..................................................................... 13 CONCLUSIONS .......................................................................................................................................... 19 STRENGTHS OF THE STUDY ....................................................................................................................... 20 LIMITATIONS OF THE STUDY ..................................................................................................................... 20 FUTURE STEPS .......................................................................................................................................... 20 REFERENCES ........................................................................................................................................... 21 TABLES ...................................................................................................................................................... 23 FIGURES .................................................................................................................................................... 33 APPENDIX 1 – LETTERS OF INFORMATION ................................................................................... 36 APPENDIX 2 – REMINDER LETTERS (ONE AND TWO)................................................................. 40 APPENDIX 3 – QUESTIONNAIRE......................................................................................................... 45

ii

List of Tables TABLE 1. QUESTIONNAIRE SCALES AND SOURCES ........................................................................................ 23 TABLE 2. RESPONSE RATES BY LOCAL.......................................................................................................... 24 TABLE 3. DEMOGRAPHIC AND JOB CHARACTERISTICS OVERALL AND BY JOB CLASSIFICATION ................... 24 TABLE 4. INCIVILITY AND SEXUAL HARASSMENT OVERALL AND BY JOB CLASSIFICATION.................... 25 TABLE 5. INCIVILITY SITUATION AND INSTIGATOR CHARACTERISTICS OVERALL AND BY RELATIONSHIP TO VICTIM................................................................................................................................................. 25 TABLE 6. SEXUAL HARASSMENT SITUATION AND INSTIGATOR CHARACTERISTICS OVERALL AND BY RELATIONSHIP TO VICTIM ................................................................................................................... 26 TABLE 7. UNION INVOLVEMENT AND ATTITUDES OVERALL AND BY JOB CLASSIFICATION .......................... 26 TABLE 8. WORST PARTS OF JOB OVERALL AND BY JOB CLASSIFICATION ..................................................... 26 TABLE 9. SELF-RATED HEALTH AND PAIN OVERALL AND BY JOB CLASSIFICATION ..................................... 27 TABLE 10. INJURIES AND SAFETY OVERALL AND BY JOB CLASSIFICATION ................................................... 28 TABLE 11. TYPES OF DISCRIMINATION OVERALL AND BY JOB CLASSIFICATION ........................................... 29 TABLE 12. CORRELATIONS ............................................................................................................................ 29 TABLE 13. SCALE DISTRIBUTIONS ................................................................................................................. 30 TABLE 14. ADJUSTED EXPLAINED VARIANCE VALUES (%) FOR MULTIPLE LINEAR REGRESSION MODELS, . 31

List of Figures FIGURE 1. ANXIETY BY SUPERVISOR INCIVILITY AND COWORKER SOCIAL SUPPORT.................................... 33 FIGURE 2. DEPRESSION BY SUPERVISOR INCIVILITY AND COWORKER SOCIAL SUPPORT .............................. 34 FIGURE 3. HOSTILITY BY SUPERVISOR INCIVILITY AND COWORKER SOCIAL SUPPORT ................................. 35

iii

Summary Background In a previous study, we conducted 60 cross-Canada focus groups with Canada Post Corporation (CPC) employees to identify health and safety issues in the workplace. It was clear from the discussions that a general lack of respect was present, and that it was linked to well-being and job satisfaction. Previous research examined these ideas by looking at incivility, defined as discourteous behaviour or treatment. Incivility is generally less obvious and is therefore probably more common than bullying or harassment. Although previous research has shown a relationship between incivility and psychological distress and well-being, incivility has not yet been explored in a Canadian population. The current study involved a questionnaire sent to workers randomly sampled from Canadian Union of Postal Workers (CUPW) locals across Canada. Study Aims The main aims of this study were: (i) to determine how common incivility was, with and without sexual harassment among Canadian postal workers; and (ii) to determine how much incivility contributed to psychological distress. Other questions were also explored and tested. Methods CUPW is the largest bargaining unit of CPC employees, representing more than 49,000 individuals. These individuals have many job responsibilities and they can be classified as ‘inside’, ‘outside’, and maintenance workers. Of the 212 CUPW locals across Canada, 12 were chosen for this study. The larger ones had been included in the previous focus groups study. The smaller ones were randomly chosen from a list of locals. Up to 225 members were sampled from each local, with members from each of the three broad job classifications. Questionnaire items were chosen based on issues that had been identified

iv

by the focus groups, and were organized into broad sections covering many topics. Since a general lack of respect was identified as a core issue, incivility was chosen as the main ‘exposure’ of interest. The Workplace Incivility Scale was used to measure the frequency of uncivil behaviours from supervisors, coworkers and customers over the previous year. The questionnaire was pilot tested with a small group of union representatives, and was then revised. Reminder letters were also mailed to increase the number of responses. Standard statistical methods were used to answer the research questions. Different measures were used to examine other characteristics like stress, depression, anxiety, hostility, job demands, and job satisfaction. Results A total of 1,968 members were sampled and sent questionnaires. Forty-two workers were omitted due to ineligibility, and 965 questionnaires were returned to the research team, making the overall response rate just over 50%. The sample was mainly male and had a mean age of 47 years. At least some incivility was quite common in the sample, with 85% of people reporting some incivility. Thirty-three percent reported incivility in the absence of sexual harassment. The rate of incivility did not significantly differ by job classification. In comparison to incivility, discrimination was uncommon in the sample. One question asked if some parts of their job were worse than others. Responses confirmed the concerns that were raised in the focus groups study. Physical strain and safety were noted most often. Fewer than 20% of the workers felt completely safe in the workplace. Overall, members most often rated their health as very good or excellent, though a surprising number (20%) rated their health as either fair or poor. Certain types of pain, specifically low back and/or buttock pain as well as neck and/or shoulder pain were experienced at least some of the time by more than half of the workers. Almost half v

of workers had been injured at work in the previous 12 months, although a surprising number of individuals did not report their injury even if it required medical attention. From the more detailed analysis, it was clear that demographic and job characteristics were not related strongly to the measures of health. Supervisor and coworker incivility did predict the health measures. Incivility was particularly strong in predicting burnout, anxiety, depression and hostility. Job satisfaction was better explained by job strain and social support than by incivility. The odds of experiencing incivility and sexual harassment decreased with age, and the odds of incivility were lower for part-time workers and temporary workers in comparison to full-time workers. Discussion As was originally found in the focus groups study, the treatment of workers by supervisors, coworkers and customers is a problem among Canadian postal workers. At least some instance of incivility was very common in the sample across all job classes. The frequency of incivility was slightly higher than in a previous study of court workers in the U.S. Incivility predicted various health outcomes after allowing other possible factors. Specifically, incivility predicted levels of anxiety, depression, burnout and hostility. This study does not necessarily prove that incivility causes higher levels of anxiety, depression, and hostility. It could be that people who are more bothered by the behaviour of their coworkers and supervisors are more anxious anyway; or there might be something else that leads to higher levels of incivility and anxiety. In other words, if this were the case, our results would just be coincidence. We think that this is unlikely, and that efforts should be made to deal with the ‘exposures’ we studied.

vi

Introduction Background In a previous study, cross-Canada focus groups were conducted with postal workers in response to concerns from both the Canadian Union of Postal Workers (CUPW) and Canada Post Corporation (CPC) (Haines, Epp, Tomosky-Chambers, Gosselin-Paiement, Shannon, & Blythe, 2006). The 60 focus groups were designed to gain information on health and safety issues in the workplace from workers at both smaller and larger worksites. The workers had a very broad understanding of health and safety in the workplace and identified issues in the areas of physical environment, psychological and physical distress, inadequate communication on policies and procedures, and relations between staff, management and workers. This information was consistent with other literature on the risks of postal work, although most research has focused on the ergonomic and musculoskeletal strain issues associated with both sorting and carrying mail (Wells et al., 1983; Ayoub and Smith, 1999; Blosnick et al., 1994; Derksen et al., 1994; Parikh et al., 1997; Stalhammar et al., 1996; Louhevaara et al., 1990). From the focus group discussions it was clear that a general lack of respect was present, and that it was linked to overall well-being and job satisfaction.

Incivility Previous research by Cortina et al. (2001) examined these ideas by looking at incivility among workers in the United States federal court system. Cortina et al found that 71% of the employees reported the experience of at least some incivility in the 5 years before the study. Incivility can be broadly defined as discourteous behaviour or treatment, and includes, for example having someone talk about you behind your back, being ignored or excluded, and condescension. The concept is distinct from bullying in that it is generally 1

less obvious. As a result, it is potentially more common than bullying or harassment because it is not necessarily overt behaviour. Incivility is postulated to be associated with several negative outcomes, and some researchers suggest that it could lead to more severe violence (Andersson and Pearson, 1999). Previous research has shown a relationship between incivility and psychological distress and well-being (Cortina et al., 2001). The current study was designed to look at incivility in Canadian postal workers, and to see how it was related to several health and psychological outcomes. Although other researchers have looked at incivility in workers, we believe this study is the first to examine incivility among postal workers.

General Approach For the current study workers were randomly sampled from 12 CPC workplaces across Canada. All sampled workers were sent questionnaires to measure multiple work ‘exposures’ (including incivility) and outcomes (e.g., burnout). The analysis looked at which exposures predicted various outcomes.

2

Study Aims As set out in the research proposal, the main aims of this study were to determine: the magnitude of incivility, with and without associated sexual harassment, among Canadian postal workers; the extent to which incivility contributed to psychological distress. In addition, secondary questions were explored: the effects of incivility on job satisfaction; the effects of sexual harassment on job satisfaction; the impact of various work organization factors (job control, job demands, social support at work) on distress and satisfaction; the demographic and job factors associated with the experience of incivility and sexual harassment; the characteristics of instigators of incivility and sexual harassment; how distinct the measure of incivility is from measures of bullying and sexual harassment. Further, we hypothesized that: incivility and bullying would only be moderately related; incivility would predict health outcomes even after allowing for work factors such as job strain; social support and coping skills would help to reduce any effects of incivility on health.

3

Methods and Results Study Population Unionized Canada Post Corporation (CPC) employees are represented by four bargaining units in four different unions. The Canadian Union of Postal Workers (CUPW) is the largest bargaining unit with more than 49,000 employees. These employees are responsible for mail processing and handling (‘inside’ workers), mail collection and distribution (‘outside’ workers), and maintenance. There are 212 CUPW locals of varying size across Canada. A total of 12 sites were chosen for the study. Nine of these sites were chosen because of their involvement in the earlier focus group study, while three sites were chosen for their small membership size. CUPW members in each of the three broad job classifications (inside, outside, maintenance) were recruited for the study.

Sampling: Who was in the study In addition to the nine sites that had participated in the earlier focus group study (Edmonton, Vancouver, Hamilton, Toronto, Welland, Sudbury, Montreal, Quebec City, St. John’s), three sites were randomly selected from the 166 locals with fewer than 100 members (Cumberland, Haut-de-Lac, Terrace). We had three Ontario locals (Hamilton, Toronto, Welland) since almost one-third of CPC employees work in Ontario. A CUPW membership list was obtained for each of the 12 locals, and workers were randomly selected, stratified by job classification. The aim was to have up to 225 workers from each local, with 100 inside, 100 outside and 25 maintenance workers. This was to ensure that there were enough workers from within each job class to allow meaningful results to be obtained for each group. In smaller locals all members were asked to take part in the study.

4

Questionnaire Item Selection: What questions were asked

Questionnaire items were chosen based on issues that had been identified by the focus groups. These included physical environment, psychological and physical distress, inadequate communication about policies and procedures, and relations between staff, management and workers. Questions were divided into broad sections covering the job, union, work environment, health and feelings, workplace situations, injuries and safety, and demographics. Since a general lack of respect was identified as a core issue, incivility was chosen as the main ‘exposure’ of interest. The Workplace Incivility Scale (WIS) measures the frequency of uncivil behaviours (disrespectful, condescending, rude) from supervisors, coworkers and customers over the previous year (Cortina et al., 2001). Questions were rated on a 5-point scale ranging from never to most of the time. We suspected that incivility would be related to several psychological and health outcomes, so information on job satisfaction, back and shoulder pain, distress, anxiety, burnout, depression, stress, and hostility was also collected. Other exposures of interest included the job content items (job control, job demands, and social support), sexual harassment, and bullying. A list of all scales used in this project, as well as their sources, is provided in Table 1. Pilot Test

The questionnaire was pilot-tested with a group of ten postal workers. The group was composed of union representatives who willingly agreed to participate. The pilot test was intended to ensure that the questions were understandable, and that the questionnaire was not too long. Respondents had the opportunity to give feedback on the questionnaire about items that were confusing, redundant or inappropriate, as well as to comment on

5

topics that they thought were missing. Following the pilot test, the questionnaire was revised and shortened. The final questionnaire was then translated into French, and backtranslated to English to ensure a proper translation. The questionnaire is shown in the Appendix.

Mailing Workers’ home addresses were obtained from CUPW and sampled workers were sent packages that included: (1) an information letter from the research team, (2) a blank questionnaire, and (3) a postage-paid return envelope. In addition, union members in good standing (i.e., those expressing interest in the union and with up-to-date union dues) were also sent a letter from CUPW indicating its support for the project. Copies of all letters are provided in the Appendix. To maximize response from the sampled workers, the ‘Dillman approach’ was used. This method involves three mail outs: a first mail out to all sampled workers, a second identical mail out with the addition of a reminder letter, and a final mail out. Since all questionnaires were anonymous we could not know which individuals had completed questionnaires after the first and second mail outs. Therefore, unless individuals contacted the research team saying that they had completed the questionnaire, or mail was returned to sender, all sampled individuals were sent the questionnaire three times. Completed questionnaires were returned directly to the research team so that the respondents could feel more comfortable replying honestly to all questions without fear of job consequences.

6

Response Rate

From the twelve locals, a total of 1,968 workers were sampled and sent questionnaires. After omitting ineligible workers because they had moved, retired, died, or gone on medical leave (42), the valid sample size was 1,926. Over three mailings, 965 questionnaires were returned, for a response rate of just over 50% (965/1,926). The response rates differed slightly by local, ranging from a low of 38% in Vancouver to a high of 71% in Terrace (see Table 2). Demographic Characteristics

The majority of the respondents were male (60%), and the mean age was 47 years (SD=81). Most were married (70%), and had worked for CPC for a mean of 18 years (SD=10). Table 3 provides a summary of the demographic and job characteristics. All demographic and job characteristics differed significantly by job classification (inside, outside, and maintenance). Maintenance workers, for example, were more likely to be male, married, and working full-time in comparison to inside and outside workers.

Data Entry All questionnaires were hand entered into a Microsoft Excel spreadsheet by two students hired for the project. The students entered responses as they were recorded on the questionnaire without extra interpretation, and also coded if the data was missing or if the field was not applicable to ensure completeness of entry. Responses to open-ended fields in French were first entered exactly as written, and then translated to English by one of the bilingual students.

1

SD=Standard Deviation (a measure of the spread of the data, small standard deviations indicate minimal variability in the sample) 7

Descriptive Analysis Scales were coded according to methods used by other authors. See Table 1 for a description of each scale used and its section in the questionnaire. For the most part, scales were calculated as mean scores of the scale items. Frequencies and/or means of demographic and/or work characteristics were obtained. This was done for all respondents combined, and for each job classification separately. Probability values (p-values) are also provided where appropriate. Sometimes, comparisons may appear to show differences or relationships, but they could have occurred just by chance. For statistical testing, p-values of less than 0.05 are generally interpreted as ‘significant’, meaning that the result would be expected less than 1 time in 20 (5%) if it were due to chance alone. Therefore, if a p-value is less than 0.05, it means that it is unlikely that the result occurred by chance. Asterisks (*) have been used to make the tables easier to interpret. The more asterisks, the less likely the results are due to chance. In other words, more asterisks mean that the evidence for the association is stronger. Incivility and Sexual Harassment

Research question: What is the magnitude of reported incivility, with and without associated sexual harassment, among Canadian postal workers? Incivility was quite common in the sample, with 85% of people experiencing at least some incivility (i.e., any incivility from supervisors, coworkers or customers), and 33% of people experiencing incivility but not sexual harassment. Only 13% of respondents experienced neither incivility nor sexual harassment. The experience of any incivility did not significantly differ by job classification; but as would be expected, customer incivility

8

was significantly more common among outside workers (55%), although it was also reported by both inside (26%) and maintenance workers (33%). Table 4 provides a summary of the experience of incivility and sexual harassment overall and by job class. Both supervisor and coworker incivility were less likely in temporary workers than in full-time workers. There were no demographic characteristics that were more common among those experiencing any type of incivility. The experience of both incivility and sexual harassment was also related to disability, defined as being on modified work (either currently or previously). Those on modified work were more likely to experience both incivility and sexual harassment (p