​South Africa’s official unemployment rate increased notably to 29.0% in the second quarter of 2019 and further to 29.1% in the third quarter. These all-time highs since the inception of the Quarterly Labour Force Survey (QLFS) by Statistics South Africa (Stats SA) in 2008 attracted attention due to their magnitude and also because significant increases historically have tended to be a first-quarter phenomenon, such as in 2009, 2015 and 2016. First-quarter increases in the unemployment rate are not unusual, as many school leavers enter the labour market (i.e. start looking for employment for the first time) at the beginning of the year, unlike in the second quarter when less seasonality is present.

In the case of the three previous first-quarter examples and in the second quarter of 2019, the notable increase in the number of unemployed people was accompanied by a decrease in the not economically active population. However, the composition of the recent second-quarter decrease in the not economically active population differs as the decreases in the previous periods resulted mostly from a decline in the student category while the decrease in the second quarter of 2019 was mostly due to a decline in the number of discouraged work seekers. In the third quarter of 2019, the number of unemployed persons increased further while the number of discouraged work seekers increased only slightly, leading to a further marginal increase in the unemployment rate.

This analysis will show that this anomaly could be ascribed to various developments. In the case of possible data and sampling methodology issues, an analysis of the QLFS micro-data shows that the characteristics of the sampled survey participants remained stable and, together with Stats SA’s rotating panel methodology, ruled out obvious sampling or methodological changes.

A plausible explanation for the sudden and sharp increase in the unemployment rate could also be the national elections that took place in the second quarter of 2019 (8 May). Temporary jobs at the Electoral Commission of South Africa (IEC), together with election campaign-related promises of job creation, could have created optimism which enticed discouraged work seekers to actively search for work. Should this have been the case, the unemployment rate might recede somewhat during the remainder of 2019 when the unemployed again become discouraged work seekers.

However, it could also be that the substantial increase in the unemployment rate, as surveyed in the second quarter of 2019, reflects the continuation of an upward trajectory over the past few years consistent with weak economic growth, subdued fixed investment and employment creation, as well as the prolonged downward phase of the business cycle.

The accuracy of the official unemployment rate is corroborated by the absence of material differences when it is compared with a counterfactual unemployment rate that hypothetically assumed no increases in the new entrants, re-entrants or ‘other’ unemployed categories while retaining the observed increases in the job losers and job leavers categories in the second and third quarters of 2019.