I’m presently coding data for a research project examining grievance arbitrations in Alberta. There are some 1000+ arbitration decisions (2006 to present) to read and code before a colleague and I can start the actual analysis. We are presently about 20% of the way through the coding. While we can't tell whether our hypotheses are correct or not (we need a much larger dataset), we do have some initial descriptive data on the 2006-2008 decisions (n=203) to share, for what it is worth.
Not surprisingly, the majority of grievance arbitration decisions come from the public sector, with the public-service, health care, and education being the most common industries. In the private sector, manufacturing, retail (grocery mostly), construction, and forestry are the industries most frequently represented.
Most arbitrations are decided by men. Grievers are about equally split between men and women, but the most numerous kind of grievances (typically policy/group grievances) tend to have mixed-gender griever groups.
Type of grievance
The three most common types of decisions about grievances address termination/discipline (30.5%), salary and benefits (22.7%), and procedural wrangling ahead of the substantive issue (15.3%). Disentangling procedural decisions from more substantive ones poses some interesting coding challenges because of how they are inconsistently reported.
Overall, employers tend to “win” most grievances (in that they achieve the outcome that they wanted). The pattern (so far) is broadly consistent with the literature.
Looking just at termination/discipline outcomes (the largest category of grievance awards), we see a similar pattern.
This is a bit surprising, because in most of these cases, the employer bears the initial onus to prove discipline/termination was warranted. This is different from most other grievances (where the union bears the initial onus). Early days though—we may see a shift as more decisions are coded.