Output, employment, the minimum wage and Covid-19

10 June 2021

INTRODUCTION

The divergence between the growth of output and the growth of employment between the first quarter of 2018 and the first quarter of 2021, as shown in Table 1 is a puzzle. An earlier brief[1] raised the possibility that the Quarterly Labour Force Survey has under-estimated employment since the lockdown in March 2020 forced a change in the way data are collected (from personal to telephone interviews, with a loss of respondents who could not be reached by telephone). This brief advances another possible reason for the divergence.

Table 1

Sources: Statistics SA Quarterly GDP estimates, Q1 2021 and Quarterly Labour Force Survey Q1 2021

OCCUPATIONS, GROSS EARNINGS AND MINIMUM WAGES

The analysis here disaggregates data by one-digit occupational category[2]. Table 2 sets out the relevant information.

Table 2

Sources: Statistics SA, Labour Market Dynamics 2019 and Quarterly Labour Force Survey, Q1 2021

Column 6 shows how unequal median monthly earnings are across occupational categories. The biggest gap is between legislators, senior officials and managers and professionals and all other occupations.

As Table 3 shows, the national minimum wage was in effect during the first quarter of 2019. It is instructive to compare gross wages as reported with the minimum wage by occupational category[3]. Some respondents refused to report their gross wages. Column 5 shows that reticence was more marked among respondents in the occupational categories with higher median wages (among those that did report). Defaults were recorded where reported gross earnings were lower than the minimum wage, and these are shown in Table 2 Column 4.

Table 3

Notice from Table 2 Column 7 and Table 3 Column 1 that, for skilled agricultural and fisheries workers, elementary occupations and domestic workers, median recorded gross earnings were lower than the minimum wage. Accordingly, the default rates in Table 2 Column 4 for these categories were above 50%. Moreover, the median gross earnings were very little above the minimum wage among service workers, shop and market sales workers and plant and machine operators and the default rates in these categories were above 40%. For the paid employed as a whole, the default rate was 40%. In occupational categories where the average wage is well above the minimum wage, the introduction of the minimum wage can be expected to have little impact on employment. The same cannot be said when the average wage is below the minimum wage, or very little above it.

Over time, one would expect default rates to fall as the system consolidates, putting upward pressure on average wage rates. We shall have to wait for the publication of Labour Market Dynamics 2020 to determine whether the default rates have in fact dropped, but the bottom row of Table 1 shows that average wages increased between Q1 2018 and Q4 2020 in both nominal and real terms. Moreover, while Table 2 Column 8 showed little change in terms of employment between Q1 2018 and Q1 2019, it showed much more change between Q1 2018 and Q1 2024, with marked drops in the five low median wage occupational categories.

INTERPRETATION AND CONCLUSIONS

The facts are in line with the hypothesis that the more thorough the implementation of minimum wage legislation, the greater will be the adverse impact on employment. There are some countervailing tendencies. Employment of professionals has expanded between Q1 2018 and Q1 2021, consistent with the relaxation of the supply constraint in this occupational category. Employment has also grown among skilled agricultural and fisheries category as a result of a couple of good years for agriculture.

On this interpretation, the effect of COVID-19 restrictions on economic activity may have been mainly to speed up an employment adjustment which would have happened anyway. The contraction of gross value added at basic prices year-on-year from Q1 2020 to Q1 2021 has made things worse.

The National Minimum Wage Act, 9 of 2018, establishes the National Minimum Wage Commission whose principal function it is to review the national minimum wage and recommend adjustments and to investigate and report annually to the Minister on the impact of the national minimum wage on the economy, collective bargaining and the reduction in income differentials[4]. The appendix to the Commission report found that:

The introduction of the national minimum wage led to a statistically significant increase but smaller than expected improvement in wages for the workers it covers. The quantitative analysis also found that it did not affect the level of employment. [Development Policy Research Unit report]

The qualitative research suggested broad compliance with the new minimum wage in agriculture, with slightly lower levels in domestic work. [Centre for Social Development Report]

Both these findings need qualification.

Although published in November 2020, the DPRU report considers only data up to Q4 2019. The report observes that the short time period during which the national minimum wage was in existence may be a limitation on their study, with some effects showing up only in the longer term, an outcome consistent with the findings in Columns 8 and 9 of Table 2.

The CSD report found that in the domestic work sector most employees reported receiving the national minimum wage with very few cases of noncompliance with the policy. This is not what was found by the QLFS in 2019, reported in Table 2 Column 4.

The possibility that the national minimum wage has had much greater employment effects than the research found must be considered. On 22 February 2021, the Minister of Labour announced that the minimum wage increases taking effect in March were the result of wide consultation and were reasonable. But the Q1 2021 official unemployment rate as measured by the QLFS was higher than at any time since the survey began in 2008. As further information becomes available, reconsideration of minimum wage policy may become necessary. Contractions of employment between 10% and 25% over three years in five occupational categories are a clear warning sign.

By Charles Simkins, Head of Research, HSF,

[1] See Charles Simkins, Has Statistics South African under-estimated the employment recovery, HSF Brief, 6 April 2021

[2] The QLFS codes information at the detailed four digit occupational category. These categories can be aggregated at the three, two and one digit levels. As the digits decrease, the categories become coarser. It is only at the one digit level that sampling error becomes small enough to establish trends.

[3] Hourly rates were used where reported hours of work were 40 or below. Monthly rates were used were reported hours of work were more than 40 hours and where hours of work were not reported.

[4] The most recent report of the Commission, published in November 2020 to recommend to the Minister a determination in 2021, can be found at https://static.pmg.org.za/201120nationalminimumwagecommission.pdf. The research behind it was contracted out to the University of Cape Town’s Development Research Unit and the University of Johannesburg’s Centre for Social Development in Africa. The research reports can be found at http://www.labour.gov.za/DocumentCenter/Publications/Basic%20Conditions%20of%20Employment/The%20Impact%20of%20the%20National%20Minimum%20Wage%20in%20South%20Africa,%20early%20quantitative%20evidence_.pdf and https://www.uj.ac.za/faculties/humanities/csda/Documents/NMW%20Report%20December%202020.pdf