Estimating the value and economic contribution of agricultural production in the former homelands of South Africa

Abstract

The value and economic contribution of agricultural production in the former homelands of South Africa has become increasingly important to measure because it is critical to our understanding of the role agriculture plays in household food security in these regions and the contribution by this section of the agricultural sector to the economy. Yet, two decades into the Democratic South Africa we still fail to consistently provide accurate estimates of this sectors value. The fundamental premise of this dissertation is to estimate the value and economic contribution of agricultural production in the former homelands of South Africa so that the subsistence agricultural sector can be well understood in terms of its characteristics and its value. The main focus of this study is therefore placed on black subsistence farmers in the former homelands of South Africa, mostly because these areas are under great pressure to maintain food self-sufficiency. The main hypothesis of this study is that, the value and economic contribution of agricultural production in the former homelands is significant when compared with the contribution by the commercial agricultural sector in South African. In order to test this hypothesis, three different data sets were analysed because none of these data sets individually provide exhaustive information for the purposes of this study. These data sets include primary data, such as the Agricultural Research Council (ARC) sample survey data from the OR Tambo District municipality conducted in 2015. The secondary data used in this study include the ARC sample survey 2013, the Income and Expenditure Survey (IES) 2010/2011 conducted by Statistics South Africa (Stats SA), and the National Income Dynamics Study (NIDS) waves 1 to 3 conducted by the Southern African Labour Development Research Unit (SALDRU). The Gross Margin (GM) analysis approach was used in this study to estimate the economic contribution of agricultural production. In interrogating the NIDS waves and IES 2010/2011 data sets, two types of variables which can be used to estimate the economic contribution of agricultural production are provided. The first type of variables are the self-reported values of agricultural goods consumed from home production, which are found in both the NIDS and IES datasets. The second type of variables are quantities of agricultural goods harvested and the value of sales from home production, found in the NIDS datasets. The variables to estimate the economic contribution of agricultural production would appear to be the self-reported values of agricultural goods consumed from home production. Using the NIDS data the estimated value of consumption from home production in current prices was R207 million based on wave 1 data, R80,5 million based on wave 2 data, and R529 million based on wave 3 data. Using the IES data the estimated value of production for home consumption in current prices was R359 million in 2010/2011. In investigating the 2010/2011 figures estimated in this study several issues arise with regard to the number of agriculturally active households and the value of agricultural goods consumed from home production. The most important issue, is that self-reported values of agricultural goods consumed by households introduce an added source of inequality to the measurement of output. According to the UNSD (2005), households can inaccurately assign values to self-produced goods because of a lack of information about local market prices. In order to avoid this source of inequality in the measurement of the agricultural sectors contribution, estimates of the economic contribution of agricultural production were pursued, based on local market prices. It was determined that only the NIDS and the ARC data sets have variables to directly estimate the economic contribution of agricultural production based on the GM approach. The variables include: quantities of crop and livestock goods harvested and the value of sales from own production. Using the ARCs data it was estimated that the annual GM per household per year in 2012 prices was R1 985.32 based on the 2013 data and R8 892.85 based on the 2015 data. Using the NIDS waves 1 and 3 data, it was estimated that the annual GM per household was R1 017.85 based on wave 1 data and R3 535.42 based on wave 3 data in 2012 prices. The NIDS wave 2 data set does not provide farm input cost and livestock production variables. As a result, it was only possible to estimate the annual Gross Farm Income (GFI) per household which was R1 973 in 2010/2011 in 2012 prices. The latter results are somewhat consistent with the ARC 2013 and 2015 figures, although not directly comparable. The Agricultural Research Council-Department of Rural Development and Land Reform (ARC-DRDLR) project introduced in the OR Tambo District municipality has played a key role in terms of changing the mind-set of farmers. Therefore, programmes such as the ARC-DRDLR project should be introduced with more vigour. Such programmes should, however, not undermine subsistence households consumption type activates.