Vusi Mathonsi was employed by Scaw Metals (Pty) Ltd as a Charge and Mentor and Assistant Administrator. In February 2015, management implemented controls requiring pre-approval for overtime work, which significantly reduced Mathonsi's weekly salary by 50%. Mathonsi became upset about this and blamed his manager, Melodie Segole, for the reduction. On 17 March 2015, Mathonsi allegedly entered Segole's office and physically assaulted him. Following a disciplinary hearing, Mathonsi was dismissed on 24 June 2015. He referred an unfair dismissal dispute to the Metal and Engineering Industries Bargaining Council (MEIBC). The arbitrator found the dismissal to be both procedurally and substantively fair. NUMSA, acting on behalf of Mathonsi, brought an application to review and set aside the arbitration award.
The review application was dismissed. There was no order as to costs.
In a review of an arbitration award, the test is whether the decision is one that a reasonable decision-maker could not reach (Sidumo test). Credibility findings made by an arbitrator should not be readily interfered with on review, as the arbitrator is in the best position to assess witnesses who appear before them. Such findings should only be set aside if entirely at odds with or completely out of kilter with the probabilities and the totality of evidence on record. A court conducting a review must consider the totality of the evidence and not adopt a piecemeal or fragmented approach analyzing each factor individually. In misconduct dismissals, once an employer establishes a prima facie case of misconduct, the evidentiary burden shifts to the employee to prove their defence; failure to do so results in the employer's prima facie proof becoming conclusive. An arbitrator's award is reasonable and should be upheld when there is a material connection between the evidence and the result, and the result is reasonably supported by evidence.
The Court observed that it would often happen that in assessing the reasonableness of an arbitration award, the court feels it would have arrived at a different decision, but in such cases the court must remind itself that the task of determining fairness is primarily given to the arbitrator and the system would not work if courts interfered with every decision simply because they would have dealt with the matter differently. The Court noted that the Sidumo test is stringent and will ensure that awards are not lightly interfered with, ensuring that CCMA awards are final and binding as long as they cannot be said to be decisions a reasonable decision-maker could not have made. The Court also commented that findings by commissioners relating to demeanour and candour of witnesses, and how they came across when giving evidence, would normally be entirely unassailable, as the reviewing court is simply not in a position to contradict such findings.
This case reinforces the application of the Sidumo reasonableness test in reviewing arbitration awards. It emphasizes that courts conducting reviews must not adopt a piecemeal approach but must consider the totality of evidence. The judgment confirms that credibility findings by arbitrators are entitled to significant deference on review and should only be interfered with when entirely at odds with the evidence and probabilities. The case illustrates the proper application of the evidentiary burden in misconduct dismissals: once an employer establishes a prima facie case, the burden shifts to the employee to prove their defence. The case demonstrates that bare denials without supporting evidence will not discharge this burden. It also reinforces that review courts must resist the temptation to substitute their own views for those of the arbitrator, and that the test is not whether the court would have decided differently, but whether the decision falls outside the range of reasonable outcomes.