On 29 August 2024, the Labour Appeal Court (LAC) delivered its judgment in an appeal by Umicore Catalyst South Africa (Pty) Ltd (Umicore) against a Labour Court ruling that had declared, among others, the retrenchment of several laboratory department employees to be substantively unfair.

Represented by the National Union of Metalworkers of South Africa (NUMSA), the employees challenged Umicore's retrenchment process, particularly the use of a ‘behavioural assessment’ as part of an overall laboratory assessment used to determine the employees to be selected for retrenchment.

The central issue in this regard was whether the behavioural assessment met the fairness and objectivity standards mandated by section 189(7) of the Labour Relations Act, 1995 (LRA). The appeal was ultimately dismissed, with the LAC finding the assessment to be overwhelmingly subjective and lacking the necessary objectivity required by the LRA.

Background

Umicore, a catalytic converter manufacturer, acquired the business of Delphi (Pty) Ltd. It initially proceeded to operate both businesses from premises leased by Delphi in Young Park as well as its own premises in Deal Party. In 2015, as part of its consolidation efforts, it decided not to renew the lease at the Young Park site and to relocate those operations to its Deal Party premises. This led to operational changes, including automation and a duplication of functions, resulting in the redundancy of 52 positions across various departments.

In line with Section 189A of the LRA, Umicore initiated a retrenchment process, holding consultation meetings that were facilitated by a commissioner of the Commission for Conciliation, Mediation and Arbitration.

Umicore used the well-known ’last in, first out’ (LIFO) principle as the primary selection criterion, except in the laboratory department, where it later introduced a ‘laboratory assessment’ midway through the process. This assessment primarily entailed a behavioural assessment in which employees were required to complete a questionnaire describing how important they viewed certain skills in the laboratory and giving examples of where they had demonstrated these skills.

Following an objection to the behavioural assessment, Umicore decided to include three additional considerations, being individual performance appraisals, disciplinary records and attendance reports for the previous two years, each of which were afforded the same weight as the behavioural assessment.

The assessment panel (panel) comprised the laboratory manager and two senior managers from other departments. Each member of the panel scored employees separately and submitted their results to the HR manager.

Aside from two senior employees, all laboratory employees were invited to participate in the assessment. Four laboratory employees, represented by NUMSA, refused to participate, arguing that the assessment was unfair and subjective. Their assessments were conducted by the panel in their absence and based on the outcome, they were selected for retrenchment and dismissed.

NUMSA, acting on behalf of the employees, challenged the dismissals in the Labour Court, contending that Umicore had acted unfairly and that the assessment was a means, opportunistically, of targeting employees with historically higher salaries for retrenchment.

Based on the evidence, the Labour Court rejected NUMSA’s allegation that the employees had been targeted purely because of their salary. However, the Court nevertheless found the dismissals to be substantively unfair, on the basis that Umicore had deviated from the agreed selection criteria by unilaterally introducing the behavioural assessment. 

Findings in the LAC

While the LAC held that the Labour Court erred in the basis for its finding that the dismissals were substantively unfair (because the selection criteria had not been agreed between the consulting parties), it upheld the outcome in the Labour Court.

In particular, the LAC found that there were several concerns with Umicore’s chosen method for selection. In this regard, the court made the following key findings:

First, by using the questionnaire as the basis for the selection, Umicore decided to focus on the employees’ personality characteristics, including initiative, enthusiasm and determination. These are inherently subjective considerations and, absent agreement, ought not to have been included in the circumstances.

Second, although past performance was considered, no individual on-the-job evaluations were conducted to assess employees' ability to work independently in the laboratory. Despite the behavioural assessment's stated objective to test for independent work and multi-tasking, most of the questions posed did not align with these goals, making the process unfair to the employees.

Third, the questions in the assessment were inherently flawed. Of the 12 questions, 10 were double-barrelled, addressing multiple issues but allowing only a single response. Some questions focused on employees' perceptions of the importance of certain skills, assessing their ability to explain the importance of these skills, rather than directly assessing their competence in performing that work. Others emphasised outward enthusiasm and personal communication styles rather than actual work results. Additionally, the framing of questions about attendance implied that employees could be penalised for taking their entitled leave, regardless of its impact on operations.

The Court further found that there is a difference between grading employees based on the opinion of their superiors and an assessment based on observations of their actual performance. In this case, not all panel members had personal experiences with the affected employees and had to rely on the laboratory manager’s opinion of the skills and attributes of the employees to help them arrive at their own scores.

Umicore relied on the case of NUMSA obo Mdleleni and Others v Faurecia Emission Control Technologies (Pty) Ltd (FECT case), in which the Labour Court found the skills assessment used by the employer to be fair and objective.

However, one of the key differences with the assessment used by the employer in the FECT case was that it was an uncontroversial, pre-existing method of skills assessment and there was no real challenge to the objective nature of the criteria. The selection method used in the FECT case also matched the operational objective, which the laboratory assessment used by Umicore did not.

Ultimately, the Court held that Umicore had failed to prove that the employees were selected based on fair and objective selection criteria and on this basis, the appeal was dismissed.

Key takeaways

This judgment highlights the necessity for employers to use fair and objective selection criteria in retrenchment decisions, where there is no agreement with the relevant consulting party on the criteria to be used. While skills may be an appropriate criterion, this judgment highlights that the fairness of this method of selection often boils down to the manner in which it is implemented.

Where questionnaires are used to assess employees’ skills, employers can expect a court to consider the granular details of such questionnaires, including for example, the topics addressed, how the questions are formulated and their weighting, in its efforts to assess fairness and objectivity. Subjective factors, like personality traits, should be avoided in these assessments. It is also advisable to consider employees’ actual, observable, work output, rather than basing assessments on the opinions of managers. 

While not the focus in this case, employers should also bear in mind that challenges may arise, too, in the application of otherwise fair and objective selection criteria. They should accordingly ensure not only that there is a fair and objective basis for assessing employees, but also that those assessments are conducted, and the results obtained and collated, fairly.

Written by Chloë Loubser, Knowledge and Learning Lawyer: Employment and Data Protection, Mbali Mnyandu, Associate, and Josua Loots, Candidate Legal Practitioner, Bowmans South Africa