The South African economy is in serious trouble. Second-quarter gross domestic product figures indicate that the economy contracted by 1.3 percent quarter on quarter, with the manufacturing and agricultural sectors reported to be in a technical recession.
It is a generally accepted truth that such lacklustre economic performance would not bode well for employment or for job creation in the affected sectors.
It was, therefore, with surprise that a number of analysts noted what appeared to be very counter intuitive labour force movements in the agricultural sector as reported by Statistics SA in the three Quarterly Labour Force Surveys (QLFS) reported on this year.
Statistical jump
If one compares the 2015 QLFS estimates to estimates reported in the corresponding quarters of last year, one sees substantial year-on-year increases in the agricultural labour force.
A rudimentary comparison of estimates shows that in the first quarter of this year, agricultural sector jobs increased by 183 000 year on year; this was followed by an increase of 200 000 jobs when comparing the second quarter of this year to the second quarter of last year, with 85 000 of these jobs having been “added” in provinces most plagued by drought and financial losses.
In the QLFS released at the end of last month, a year-on-year comparison shows the sector managed to create 211 000 jobs. In a sector that has been ravaged by drought and whose output has shrunk significantly as a result, such employment creation should certainly seem odd.
Yet, following the release of every QLFS this year, the “outstanding” employment performance of the agricultural sector has been lauded by many analysts and headlines on this matter make it appear that we were witnessing a small jobs miracle in agriculture, despite the drought.
While such conclusions are a ray of hope in an otherwise grim national economic situation, they are unfortunately inaccurate and based on a misinterpretation of Stats SA’s data and a lack of understanding of Stats SA’s sampling methodology.
Following the 2011 census, Stats SA introduced a new master sample to account for the changes in the structure of the underlying population, which had changed significantly since the previous census.
This new master sample is now used as the basis for surveys such as the QLFS. Stats SA indicated that from the first quarter of this year, QLFS estimates had been based on the new master sample, while employment estimates in previous quarters were based on the old master sample.
The year-on-year changes this year have thus mostly been influenced by the change in the updated sample and not actual labour market movements. Data reported in the 2015 QLFSs is not entirely comparable to data reported in 2014 QLFSs.
By drawing comparisons, one would not be comparing apples with apples and thus risks making false conclusions, as has often been the case with agricultural jobs. While it is impossible to precisely quantify the changes in the agricultural labour force, what we do know for sure is that the sector did not miraculously create 150 000 jobs at the start of this year.
We also know that agricultural employment has been on a downward trend since the 1970s and no major structural changes have occurred to reverse this trend.
It should be noted that a revision of the figures must not be interpreted to mean that the old, pre-2015 estimates of employment produced by Stats SA were wrong; determining the size of South Africa’s labour force can never be an exact science.
Only time will tell whether the new higher estimates of agricultural employment are the “new normal”, and until Stats SA has had sufficient time to monitor the implications of the master sample revision, we won’t know for sure.
For now, what is clear is that year-on-year comparisons in relation to the QLFS should be approached with caution and analysed in the context of prevailing economic realities.
Rosy picture
Reporting that inaccurately paints a rosy picture in relation to agricultural employment at a time when the sector is taking serious strain not only sends confusing messages to investors, it also diminishes the plight of thousands of farmers and farmworkers whose livelihoods have been severely affected by the drought.
In line with international best practice, Stats SA is entitled to revise its samples when it deems fit to improve the efficiency of its estimates. This is recognised.
The example of the agriculture sector shows, however, that Stats SA’s sampling methodology is not fully understood by the main users of its data, thus creating room for misinterpretation and false reporting.
Given the importance of employment statistics for policy, investment decisions and business strategy, better communication by Stats SA regarding how to interpret the data it produces is definitely in order.
Without a clear understanding by all about what QLFS data means and how it should be interpreted, the false belief that a jobs miracle has occurred in the agricultural sector will continue to thrive.
* Thabi Nkosi is the senior economist at Agri SA. Follow her on Twitter at @MissThabiNkosi.
** The views expressed here are not necessarily those of Independent Media.
BUSINESS REPORT