# Annex A. Methodological annex

The report provides an in-depth view on a diverse set of countries, which vary in their income levels, development, size and importance of the agricultural sector (as a percentage of gross domestic product, GDP), and business environment. Five of the countries are in sub-Saharan Africa and two are in Southeast Asia. Table A A.1 lists the sample countries and provides summary statistics for each, highlighting the report sample’s heterogeneity. These countries represent a diverse set of agri-food producers, with varying levels of development, urbanisation and labour market dynamics.

South Africa

South Africa is an upper middle-income country with an economy largely based on services, manufacturing and mining (FAO, 2016[5]) . While agriculture contributed only 1.9% of GDP and 5% of employment in 2019, South Africa has a highly developed and diversified agricultural sector, with climate conditions that are conducive to growing a variety of crops (World Bank, 2021[1]; US International Trade Administration, 2020[6]). Main commodities include maize, wheat, sugarcane, sunflower, potatoes, groundnuts, citrus and grapes. Close to half of the country’s agricultural production value is derived from animal products. South Africa is a net exporter of food and primary agricultural products and imports processed foods. The agriculture sector is generally characterised by commercial farmers and subsistence smallholders engaging in intensive crop production and mixed farming, as well as cattle-ranching and sheep-farming (FAO, 2016[5]). The food economy in South Africa is diversified, with food manufacturing and services segments especially well developed, namely in the food retail sector, mostly dominated by large supermarket chains. The value added in services represents nearly 61.2% of GDP and 72.4% of employment (World Bank, 2021[1]). Recent years have seen the rise of large agri-businesses and agro-processors in the food systems value chain, with a decline in mid-sized farms and farmers increasingly engaged in contracts with supermarkets or processors at the expense of alternative market sources (Battersby, Marshak and Mngqibisa, 2016[7]).

Namibia

Namibia is an upper middle-income country. Agriculture is notoriously difficult in Namibia, as the country’s climate is extremely arid. In 2019, agriculture represented 6.2% of the country’s GDP and 21.8% of employment (World Bank, 2021[1]). Livestock farming constitutes approximately two-thirds of agricultural production and 24.5% of employment in agriculture, with crop farming and forestry making up the remaining production (US International Trade Administration, 2020[8]). Livestock also contributes to the majority of Namibia’s exports by value, also around two-thirds, while the export value of crops has been rising in recent years (US International Trade Administration, 2021[9])Services are the country’s main contributor to GDP as well as source of employment, with the value added in services representing 58.3% of GDP and the sector employing 61.7% of workers in 2019 (World Bank, 2021[1]).

Zambia

Zambia is a lower middle-income country. In 2015, Zambia’s agricultural sector contributed only 5% to GDP but employed 51.7% of its workforce (World Bank, 2021[1]). Within the food economy, 89% of jobs are in agriculture, followed far behind by food trade (7.5%). Smallholder farms dominate the agricultural sector in Zambia, but a number of medium- and large-scale farms produce cash crops and other food items for domestic and export markets (Jayne et al., 2019[10]; US International Trade Administration, 2020[11]). The services economy, on the other hand, contributed 56.2% to GDP and employed 40% of the workforce in 2019 (World Bank, 2019[12]). Nutrient mining, soil erosion and poor water management continue to be the main constraints for further agricultural development and increased yields. It is estimated that, by 2025, maize production will decrease due to climate change, while the variability of yields may increase (FAO, 2021[13]).

Tanzania

Tanzania is a low-income country. Tanzania is heavily dependent on the agriculture sector which, in 2019, contributed 25.8% to its GDP and employed around 65% of its workers (World Bank, 2021[1]). The country’s top commodities are primarily comprised of agricultural products, including cassava, maize, sweet potatoes and sugar cane (FAOSTAT, 2021[14]). Tanzanian farmers are shifting from smallholder and subsistence farming towards medium-scale and export-oriented production, and urban farming is increasing as well (Wineman et al., 2020[15]). Yet, productivity and the share of cultivated over arable land remain low, and farmers continue to rely on rain-fed agriculture and make use of limited modernised farming techniques (FAO, 2021[16]; WFP, 2021[17]). Tanzania has the third largest livestock population in Africa, as well as extensive natural resources for livestock development, including diverse vegetation and vast rangelands. In spite of these resources, the livestock sector is underperforming, as it only contributes 7.4% of the country’s GDP and grows at a rate of 2.2% annually. Main constraints to growth include low livestock reproductive rates, high mortality and high disease prevalence. The services sector is growing, accounting for around 41.3% of GDP and approximately 28% of employment in 2019 (World Bank, 2021[1]). Moreover, the recent discovery of large natural gas and oil reserves presents the opportunity for new and significant sources of revenue for the country (WFP, 2021[17]).

Uganda

Viet Nam

Viet Nam is a lower middle-income country with a rapidly developing economy, diminishing reliance on export-led agriculture, and rising economic growth due to industrialisation and services. Agriculture in Viet Nam benefitted from a Green Revolution, which saw advances in agricultural technologies that boosted factor productivity, such as irrigation techniques, seed varieties, fertiliser and pest control (Hazell, 2009[22]). Agriculture contributed to 16.3% of GDP and 37% of employment in 2019 (World Bank, 2021[1]), and the main commodities produced include rice, vegetables, sugar cane, cassava, maize, meat, fruit, bananas and coffee (FAOSTAT, 2019[23]). Commodities such as rice continue to be grown mainly by smallholders for local consumption, with only around 5% exported, while growing urban food markers have put increasing pressure on food production systems to modernise (GIZ, 2021[24]). Food trade is still an important feature in Viet Nam’s food economy, providing about 13.4% of food economy jobs. In 2019, food manufacturing comprised 3.8% of food economy jobs in Viet Nam whereas food service jobs made up 4.9% of the country’s food economy jobs. In general, services made up 40.9% of the country’s GDP and 35% of its labour force (World Bank, 2021[1]).

Thailand

Thailand is an upper middle-income country. Agriculture only comprised 8.4% of Thailand’s GDP and 31% of its total employment in 2019 (World Bank, 2021[1]). Food manufacturing comprised 6.2% of Thai food economy jobs, whereas food service jobs made up 10.5%. Similar to Viet Nam, agriculture in Thailand benefited from a Green Revolution which modernised its agricultural practices (Hazell, 2009[22]). Sixty-eight percent of land for field crops is arable, and main commodities consist predominantly of sugar cane, followed by cassava, rice, palm oil, rubber, maize, fruit, meat and pineapples (FAOSTAT, 2019[25]; ITC, 2021[26]). Thailand is the largest exporter of tapioca products, rubber, frozen shrimp, canned tuna and canned pineapple in the world (US International Trade Administration, 2021[27]) and the agriculture sector largely consists of small-scale, family-owned and family-operated farms (FAOSTAT, 2019[25]). However, the agriculture sector generates the lowest value added per worker and exhibits the slowest growth relative to other economic sectors (UN Thailand, 2020[28]); it has largely been overtaken by services, Thailand’s main sector of economic activity, which makes up 56.5% of its GDP and 46% of its workforce (World Bank, 2021[1]).

Household surveys were used to produce the analysis in Chapter 1, with the exception of South Africa where a labour force survey was used. Table A A.2 lists the data sources for each sample country. Unless stated otherwise, figures provided throughout the chapter for each country correspond to the year of the survey in question.

Employment in the food economy is assigned using the information from the United Nations International Standard Industrial Classification of All Economic Activities (ISIC) that accompanies each job. Following the classification scheme outlined in Allen et al. (2016[29]) and Allen, Heinrigs and Heo (2018[30]), the food economy has been grouped into the following four broad categories:

1. 1. Food agriculture: activities within the primary sector dedicated to the production of agricultural and animal products for human consumption (ISIC divisions 1, 3)

2. 2. Food processing: activities related to processing and manufacturing food and beverages for human consumption (ISIC divisions 10, 11, 12)

3. 3. Food marketing: all transport, wholesale and retail activities related to food (ISIC codes 4653, 4711, 472, 4781).

4. 4. Food-away-from-home: restaurants, street food and other catering services (ISIC division 56).

As the ISIC categorisation scheme does not sufficiently disaggregate between the types of cargo transported, isolating food-related transport was not possible. An adjustment was made for the estimates on food-related transportation, using aggregate food expenditure as a proxy for the aggregate food demand of the country in question. Estimates of aggregate food demand were made using a national average of food consumption in total consumption taken from household surveys in the same year for each sample country.

For certain parts of the analysis, a full-time equivalent (FTE) method of defining jobs was used. This method defines one FTE as 40 hours worked per week and estimates employment by converting the total number of hours worked into the number of FTE jobs (that are equal to one full-time equivalent in employment). Full-time employment is defined over the course of one calendar year (52 weeks). Full-time equivalents were generated based on the number of hours worked per week per job, over the course of 52 weeks.

“Youth” refers to young women and young men between the ages of 15 and 29. “Adults” refers to all those aged above 29. “Working-age population” refers to all those aged above 15.

Chapter 2 sought to contribute to our understanding of the employment potential in the food economy by forecasting the changes in employment directly associated with rapid urbanisation and the rise of the middle class, holding all else constant. The projections show that such trends would increase the overall level of employment in the food economy, in absolute terms in the case of agriculture and, in the case of downstream segments, in both absolute and relative terms. Looking at the changes within the different segments of the food economy further shows a rebalancing of food economy employment from the agricultural sector to secondary and tertiary food economy activities.

All in all, these results suggest that if local food systems were to take on the challenge of responding to higher and changing domestic demand for food, a large number of new jobs could be created in the food economy. Chapter 3 discussed how different types of local food systems could further influence the quantity and quality of employment in the food economy, while responding to the social, economic and environmental challenges.

This forecasting exercise is based on two novel, uniquely disaggregated sectoral employment datasets provided by the International Labour Organization (ILO): Employment by sex and economic activity (ILO modelled estimates) and Employment by sex and economic activity (ISIC level 2).

A two-step methodology was specifically developed to harness information from both datasets, using GDP and urbanisation as main predictors and deriving country-specific elasticities. Employment estimates at horizon 2030 relied on the United Nations’ (UN) 2018 Revision of World Urbanization Prospects (available at horizon 2030) and the International Monetary Fund’s (IMF) World Economic Outlook GDP forecasts (available at horizon 2025, extended to 2030; see Annex A for details).

Employment-growth and employment-urban population partial elasticities are the first set of parameters used to produce employment projections for the food economy. Employment elasticities measure the employment “intensity” or sensitivity of, in our case, economic growth and urbanisation. They provide important information about labour markets and serve as an indicator of how growth in economic output and growth in employment evolve together over time.

In a first step, country-specific long-term elasticities are estimated for four broad sectors of the economy:

• Agriculture; forestry and fishing (ISIC 4.0, letter A)

• Manufacturing (ISIC 4.0, letter C)

• Wholesale and retail trade; repair of motor vehicles and motorcycles (ISIC 4.0, letter G)

• Accommodation and food service activities (ISIC 4.0, letter I).

Following the IMF guidelines for forecasting labour market indicators (Chami, 2012[31]), we estimate the following for each country separately:

where ${empl}_{i,s,t}$ is the employment of sector s in country i at time t.

$\mathrm{X}$ is a matrix of explanatory variables, .

The specification is estimated separately for each country/sector over the 1990-2019 period, therefore obtaining country/sector specific parameters ${\beta }_{i,s}$, the elasticity of employment of sector $s$ in $i$ with respect to GDP and urban population (absolute level). These regressors are mainly to be interpreted as “demand” factors, as GDP growth and urbanisation are key determinants of an increase in food demand and related labour demand. Yet, they also pick up labour market supply, as urban population level is correlated with overall population level, for example. The estimation uses robust standard errors.

Employment data are obtained from the ILO’s modelled estimates, providing information on sector-level employment for all countries. GDP is obtained from the World Bank’s World Development Indicators, and urbanisation is obtained from the UN’s 2018 Revision of World Urbanization Prospects.

In a second step, a relationship between the broad sectors and the specific food economy sectors is estimated in a panel setting, exploiting the highly disaggregated Employment by sex and economic activity (ISIC level 2), recently released by the ILO. The following specification is applied:

where ${empl}_{i,s,t}^{SPECIFIC}$ is the specific two-digit sector, for example Restaurants (ISIC 4.0 division 56), ${empl}_{i,s,t}^{BROAD}$ is the broad sector containing the specific sector, in this case Accommodation and food service activities (ISIC 4.0, letter I). ${\beta }_{r,s}$ corresponds to region X income group specific parameters, estimated for each sector s. The estimation clusters standard errors at the country level.

The projected employment growth is given by the following expression:

Where X is the set of step-1 regressors.

The baseline scenarios are built around extensions of the IMF’s World Economic Outlook (October 2020) forecasts, with the baseline scenario assuming that countries return to their long-term, pre-COVID growth path.

We articulate scenarios around the World Economic Outlook forecast (IMF, 2020[32]) October update, which provides growth forecasts at horizon 2025.

For our baseline scenario, we assume that countries will grow until 2025 at the rates forecasted by the IMF. Beyond 2025, we assume that countries will return to a pre-pandemic “long-term growth path”. We define the long-term path as the average growth between 2014 and 2024 as forecasted by the IMF in 2019 (IMF, 2020[32]).

## References

[29] Allen, A. et al. (2016), Agrifood Youth Employment and Engagement Study (AgYEES), Michigan State University, East Lansing, Michigan.

[30] Allen, T., P. Heinrigs and I. Heo (2018), “Agriculture, Food and Jobs in West Africa”, West African Papers, No. 14, OECD Publishing, Paris, https://dx.doi.org/10.1787/dc152bc0-en.

[19] Anderson, J., C. Learch and S. Gardner (2016), National Survey and Segmentation of Smallholder Households in Uganda: Understanding Their Demand for Financial, Agricultural, and Digital Solutions.

[21] Atube, F. et al. (2021), “Determinants of smallholder farmers’ adaptation strategies to the effects of climate change: Evidence from northern Uganda”, Agriculture & Food Security, Vol. 10/6, pp. 1-14, http://dx.doi.org/10.1186/s40066-020-00279-1.

[7] Battersby, J., M. Marshak and N. Mngqibisa (2016), No. 05: Mapping the Informal Food Economy in Cape Town, South Africa, http://scholars.wlu.ca/hcp (accessed on 15 January 2019).

[31] Chami, R. (2012), “A Template for Analyzing and Projecting Labor Market Indicators”, Technical Notes and Manuals, Vol. 12/01, http://dx.doi.org/10.5089/9781475522051.005.

[13] FAO (2021), Economics and Policy Innovations for Climate-Smart Agriculture- Zambia, http://www.fao.org/climatechange/epic/projects/countries/zambia/en/ (accessed on 5 July 2021).

[16] FAO (2021), FAO in Tanzania, http://www.fao.org/tanzania/fao-in-tanzania/tanzania-at-a-glance/en/ (accessed on 28 June 2021).

[5] FAO (2016), Country Profile- South Africa, FAO, Rome, Italy, http://dx.doi.org/10.21552/cclr/2010/4/146.

[14] FAOSTAT (2021), Commodities By Country: Tanzania, http://www.fao.org/faostat/en/#rankings/commodities_by_country.

[25] FAOSTAT (2019), Commodities by country - Thailand, http://faostat.fao.org/site/339/default.aspx.

[23] FAOSTAT (2019), Commodities by country - Viet Nam, http://faostat.fao.org/site/339/default.aspx.

[24] GIZ (2021), Better Rice , Better Life.

[22] Hazell, P. (2009), The Asian Green Revolution, International Food Policy Research Institute (IFPRI), Washington, DC, http://www.ifpri.org/millionsfed (accessed on 3 May 2021).

[4] IFPRI (2020), Agricultural Total Factor Productivity (TFP), 2000-2016, International Food Policy Research Institute, Washington, DC.

[32] IMF (2020), World Economic Outlook, October 2020, International Monetary Fund Research Department, Washington, DC, http://dx.doi.org/10.5089/9781513556055.081.

[26] ITC (2021), Country profile - Thailand, https://www.intracen.org/exporters/organic-products/country-focus/Country-Profile-Thailand/ (accessed on 5 July 2021).

[10] Jayne, T. et al. (2019), “Are medium-scale farms driving agricultural transformation in sub-Saharan Africa?”, Agricultural Economics, Vol. 50, pp. 75-95, http://dx.doi.org/10.1111/agec.12535.

[3] UN DESA (2019), 2019 Revision of World Population Prospects, United Nations Department of Economic and Social Affairs, New York, https://population.un.org/wpp/ (accessed on 12 February 2021).

[28] UN Thailand (2020), Thai Agricultural Sector: From Problems to Solutions, https://thailand.un.org/en/103307-thai-agricultural-sector-problems-solutions (accessed on 5 July 2021).

[2] UNDP (2020), Human Development Index (HDI), United Nations Development Programme, http://hdr.undp.org/en/content/human-development-index-hdi (accessed on 12 February 2021).

[9] US International Trade Administration (2021), Namibia - Country Commercial Guide, https://www.trade.gov/country-commercial-guides/namibia-agricultural-sector (accessed on 29 June 2021).

[6] US International Trade Administration (2020), South Africa - Country Commercial Guide - Agricultural Sector, International Trade Administration, Department of Commerce, United States, https://www.trade.gov/knowledge-product/south-africa-agricultural-sector (accessed on 30 April 2021).

[11] US International Trade Administration (2020), Zambia - Country Commercial Guide - Agriculture Sector, International Trade Administration, Department of Commerce, United States, https://www.trade.gov/country-commercial-guides/zambia-agriculture (accessed on 2 May 2021).

[15] Wineman, A. et al. (2020), “The changing face of agriculture in Tanzania: Indicators of transformation”, Development Policy Review, Vol. 38/6, pp. 685-709, http://dx.doi.org/10.1111/dpr.12491.

[1] World Bank (2021), World Development Indicators (WDI), World Bank, Washington, DC, https://datacatalog.worldbank.org/dataset/world-development-indicators (accessed on 6 December 2018).

[33] World Bank (2020), World Development Indicators, The World Bank Group.

[12] World Bank (2019), World Development Indicators, The World Bank Group.

[20] World Bank (2018), Closing the Potential-Performance Divide in Ugandan Agriculture, World Bank, Washington, DC, http://www.worldbank.org (accessed on 2 May 2021).