Each LMIP Briefing is an evidence-based contribution to informing the development of a skills planning mechanism for South Africa. Briefings showcase our cutting edge research and aim to highlight key trends and potential implications from LMIP research projects.

LMIP Briefing 16

The SETA Labour Market Survey

The overarching purpose of the SETA Labour Market Survey is to provide a new tool that is missing from the current skills planning mechanism in South Africa. It does so through contributing a new methodology and dataset that provides key labour market information. A pilot survey was conducted within the Manufacturing and Engineering SETA (merSETA) during the last quarter of 2014.

The survey is designed to collect baseline data about the employment profile of each firm (including employees’ current qualifications, job title, and wage level) as well as the nature of the education and training provided (including the type and name of the programme, and its duration). Critically, the data is collected per employee, which allows for more meaningful insights. The survey collects firm-level information on the rationale for training, factors that have enabled and hindered training, training expenditure, and the level of engagement with the SETA.

Key results

- The merSETA labour market is a semi-skills intensive labour market as 45 percent of workers have a Grade 12 completion only. This suggests that training at the workplace is a key component of skills development in these sub-sectors.

- Just over a third of all employees in this labour market underwent training within the year. Those in the lowest three occupational levels (OFO groups 6-8) made up just under half of all the employees trained. Sub-sectors such as Motor and Plastics are relatively more training intensive than others such as tyre manufacturing.

- There is significant variation in the type of training provided by firms in different sub-sectors and of different sizes. The data obtained uncovers the heterogeneity in the skills needs of different types of firms (and different sized firms within the same sub-sector).

- Different types of training yield different returns to the employee and depend on the individual’s initial level of education. Returns to ‘skills programmes’ increase with the initial level of education – those with a degree who complete a skills programme have very large positive returns, whereas trainees with a matric have smaller positive returns. In addition, only at some sufficiently high enough level of initial education (above post-matric diploma), there are positive returns to ‘on the job’ training. Lastly, those with an initial level of education of at least a post-matric diploma, experience positive returns to a wider range of training programmes, and returns that are also higher.

- These results would seem to support the notion that general and specific skills are complementary and that a higher level of general education increases the returns to more specific types of training.  


1. SETA labour market surveys are a critical instrument for workplace skills planning.  Without these, we will continue to rely on the QLFS data, which remains inadequate for understanding skills dynamics at the firm level. 

2. Run enterprise surveys, specifically SETA labour market surveys, by legislating SETAs to run these surveys bi-annually, in order to provide such information to the proposed Skills Planning Unit. 

3. Individual unit record data should be the aim of such a survey, as it allows for the tracking of training of all workers over time. Only with unit record data can we measure, for example, the impact of training on employees’ labour market outcomes (e.g. wages). The flexibility of the unit record data means that various types of reports can be generated from this data – from detailed employee-level analysis to more aggregated firm and sector analysis.

4. Ensure that SIC and OFO codes used by SETAs, StatsSA and international surveys are aligned.