Estimates indicate that for Africa south of the Sahara, DVCs supply about three quarters of national food consumption and in India close to 90%. For Africa south of the Sahara, we calculated the share of imports from 2017 FAOSTAT data as the the share of imports in tons in consumption by disappearance (CbD) in tons. Using only data on output less exports plus imports as the measure of CbD, we get the share of imports at 10%. The complement is 90% which is supplied by national output. An alternate measure that we did not use here calculates CbD as output less exports plus imports, less FAO estimates per country and crop of losses, use of output as seed and feed, and use of output for non-food products, which produces a lower CbD and 13% as the share of imports in total food consumption. We did not consider stock changes because FAOSTAT signaled those as rife with missing data.
We then divided that domestic output pie into purchases, 80%, and its complement, home production, at 20%. The DVC share in total consumption nationally is thus roughly 90%*80% or 72%. The approximate figure of 80% of purchases (and thus FSCs) in total national consumption for Sub-Saharan Africa was derived as an average over LSMS data for 2010 from Malawi, Nigeria, Tanzania, Uganda, Zambia. It is reported in Reardon et al. 2019b and draws from Dolislager (2017 and 2019; Sauer et al. 2019). First, using true population weights, we derived the share of total urban consumption in national consumption. We analyzed the urban consumption module data and found that nearly 100% of urban consumption came from purchases. We found on average that 60% of national consumption came from urban consumers and thus from purchases. That left 40% of national consumption, We analyzed rural household consumption data from the LSMS; we valued farm output at the consumer purchase price based on the median revealed price during the year in the Enumeration Area of the household.
For India, we calculated the share of imports in total consumption by disappearance using FAOSTAT for 2017, all food, in tons using the CbD measure as output less exports plus imports. The share is 3.0%, hence the domestic output is 97%. Note that if we measure CbD by output less exports plus imports less FAO estimates of feed, losses, seed use, and non-food use, the import share becomes 3.2%.
We calculate the share of DVCs in domestic consumption by multiplying the domestic output share in consumption, 97%, by the share of purchases in consumption, 92%, to get 89%. For the purchase share of 92% of all consumption, we drew from Reardon et al. 2020 (EPW, pages 18-19) which states: “We obtained 92% as follows: (i) Of all the food consumed in India, 60% is consumed by urban consumers. This is calculated from the National Sample Survey Office’s (NSSO) 2011–12 data on food consumption per capita in urban and rural areas, in value terms. Essentially, all the food consumed in urban areas is purchased, since almost all urban households are net buyers of food. Note that this exceeds the urban population share because the level of food expenditure per capita is higher in the urban than in the rural areas. There are various estimates of the urban population share in India. The official one is 31% based on the 2011 Census. Onda et al (2019) estimate the share at 43% using a community-detection clustering algorithm to construct urban agglomerations for all of India. Whatever the estimate be, it is below the urban share in the overall food economy (purchases plus own production). (ii) Of the 40% of India’s food that is consumed in rural areas, 80% (in value terms) is purchased (while the rest is home-produced on own farms). This 80% figure is a conservative estimate, which in turn is a simple average1 of two estimates: the first is 84% in rural areas, based on NSSO 2011–12 (GOI 2013); and the second is 75% from the survey villages of the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), which tend to be in the “poorer than average” semi-arid zones (Kumar et al 2019). The high share of purchases in overall food consumption in these poor villages shows the high dependency of even the hinterlands on the food markets.”