How many commercial farms are there really in South Africa and are most of them ‘large’?
This article was first published in the Landbou Weekblad.
Good policy-making requires reliable, comparable statistics over time. Despite there being an annual survey of agricultural firms in South Africa, confusion exists about the number of commercial farms in South Africa and the structure of the agricultural sector. According to the agricultural census in 2007, there were 39 966 commercial farms, while the agricultural survey mention a figure of 64 192 and 57 126 in 2008 and 2017, respectively. With such diverging numbers across time, which figures should we trust and how does one analyse trends in the sector and make evidence-based decisions?
Although we do not have all the answers now, we are able to provide some clarity on the sector. We do so by using a new, anonymised administrative data set at the National Treasury containing the universe of VAT-paying entities. We have accessed this data through the Southern Africa – Towards Inclusive Economic Development programme (SA-TIED); a programme between UNU-WIDER, the National Treasury, IFPRI, and various other departments in the economic cluster, as well as a number of local research organisations.
What this dataset shows, is that there are approximately 40 000 active VAT-registered agricultural entities across years, with a slightly higher number for 2015. The number of VAT-paying entities is very constant across time, suggesting that the size of the agricultural sector likely remained the same across 2009-2016. Furthermore, our figures are similar to the figure of 39 966 as per the 2007 Agricultural census.
The most common firm size category is micro; roughly half of the entities are classified as micro while medium and large entities jointly constitute less than 20% of entities. This, therefore, provides evidence that not all commercial farms are large in South Africa. However, it does seem as if the proportion of firms classified as micro is decreasing over time while the proportion of firms classified as large is increasing.
|Turnover thresholds in 2016 Rands|
|<1 mil.||>1 mil. & <6 mil.||>6 mil. & < 10 mil.||>10 mil.|
|2009||21 106||13 306||2 477||3 814||40 703|
|2010||21 570||13 234||2 417||3 655||40 876|
|2011||21 173||13 444||2 448||3 773||40 838|
|2012||20 608||13 497||2 593||4 197||40 895|
|2013||20 483||13 210||2 655||4 524||40 872|
|2014||20 477||13 180||2 667||4 805||41 129|
|2015||17 761||14 583||3 116||6 860||42 320|
|2016||19 701||13 220||2 676||5 135||40 732|
Our results have shown that: i) the sector has not changed all too much from 2009-2016; ii) most farms fall into the micro firm size category. However, while this dataset is able to give us an indication of VAT-registered entities, many entities would not be captured in the VAT records if their annual turnover is below the compulsory VAT registering threshold of R1 million. Therefore, any survey or census which uses the VAT records as a sampling frame will automatically exclude entities which are non-VAT registered and the number of agricultural entities will be underestimated.
While the 2011 population census included questions about agricultural households, no distinction was made whether these households are involved in subsistence or commercial farming. The figure of 2.9 million households which are involved in agricultural activities is, therefore, not very indicative of the number of commercial farms. However, it does suggest that many households are involved in agricultural activities, and that many agricultural entities may not be captured in the VAT records.
To get a better estimate of the number of commercial farms, one could add the number of agricultural entities which are registered for company income tax and not for VAT to the number of VAT-registered entities. However, from the VAT records, we can see that over 60% of agricultural entities are sole proprietors while roughly 30% are closed corporations and companies. Therefore, even adding those figures will underestimate the number of commercial farms in South Africa.
With the recent release of the agricultural census results, where the VAT-registered entities formed the basis of the sample frame, one needs to remember that these results are only representative of entities which have a turnover above R1 million and that many entities are excluded from the census. The results are, therefore, likely to underestimate the number of commercial agricultural entities and only be representative of entities whose turnover is above R1 million.
We propose that the government should investigate the number of commercial farms so that i) one can update the framework on which the censuses and annual agricultural surveys are based ii) or at least understand how representative the results from the censuses and agricultural surveys are. This can clarify many misperceptions of the agricultural sector and ensure that policy-makers have access to more reliable statistics on which they can make evidence-based decisions going forward.
The sub-industries which are included in this definition are: breeding of non-food producing animals, crop farming, flower and seed growing, fruit farming, livestock farming, mixed farming, poultry farming and the production of milk.
 The thresholds used for the firm sizes are based on the figures from the National Small Business Amendment Act (No. 26 of 2003) and are adjusted for inflation.