Media statement: Postponement of the release of Quarterly Employment Statistics

MEDIA STATEMENT                                                                                                                                                     2 April 2017

Postponement of Quarterly Employment Survey – QES December 2016 release

The Quarterly Employment Survey is establishment based and covers formal sector businesses which are registered for VAT and government departments at all levels. This implies that the survey depends on the business register as a sampling frame. However, new businesses open (birth) while some businesses close (death) all the time. Therefore, it is imperative that the business register is updated continuously.  Statistics South Africa updates the frame continuously but it is not practical to draw new samples every quarter thus new samples are drawn once a year.

Number of establishments in 2015 and 2016


The table above shows the number of businesses which were eligible for inclusion in the QES sample in 2015 and 2016. The table shows that in 2016 there were 7 155 more businesses compared to 2015. New samples are introduced in June every year, however it is good practice to conduct parallel collection of data in both the new and the old sample in order to determine the impact of introducing the new sample.


Figure 1: Employment trend to September 2016 using 2015 and 2016 sample


Figure 1 shows employment estimates from both the 2015 and 2016 samples for June and September 2016. Parallel collections were conducted in June and September. The new sample generates higher employment compared to the old sample reflecting the growth in the business register. However new businesses start gradually over a period of time but not as a once off event. This has to be reflected in the employment trends. However, there is a lag between growth or declines in the business register and the drawing of new samples. Therefore to reflect the dynamics of birth and deaths of businesses, employment figures reported during the lag periods have to be revised otherwise there will be a break in the series when a new sample is introduced.

On several occasions the QES series was broken due to introduction of new samples and no revisions of the previous trends were provided.  Statistics South Africa would like to release the QES December 2016 figures using the new sample. For the first time Statistics South Africa would like to provide revised figures of previous trends with the December 2016 figures in order to avoid a break in series. However, to do so, adjustment methods have to be robust and transparent. Statistics South Africa will therefore postpone the release to 11th of April 2017 in order to go through consultations with key stakeholders on the robustness of the methods.

Issued by Statistics South Africa

Dr Pali Lehohla, Statistician-General of South Africa


For media enquiries contact:

Trevor Oosterwyk

Head of Communications

Tel: 012 310 4687

Cell: 082 908 9104