Food accessibility of different socioeconomic groups in sub-Saharan African cities: a mixed-method analysis in Kampala, Uganda

Abstract

Cities in sub-Saharan Africa are characterised by rapid urban sprawl, which has implications for urban food accessibility. Urban sprawl results in inefficient structures of cities, and is often related to patterns of socioeconomic segregation. An important research gap in food accessibility studies is that these local socioeconomic imbalances are not considered in broad-scale studies. This research analyses how the dimensions of food access (physical, social and economic) relate to the food insecurity and dietary diversity of inhabitants of different socioeconomic groups in the rapidly growing Greater Kampala Metropolitan Area (Uganda). We use the Food Insecurity Experience Scale and Household Dietary Diversity Score to assess the overall state of food consumption. To measure physical accessibility, we geographically map the formal food system potential. A radar chart was used to visualise the vulnerability of different socioeconomic groups within the city food system. The results show that more established urban dwellers experience different access vulnerabilities than newly migrated residents, depending on their income. Lower income groups compensate their limited economic accessibility by participating in food sharing networks. Obtaining a better understanding of the dimensions of urban food accessibility can aid stakeholders in the urban food system in their policy making processes towards a more food secure and sustainable future.

Appendices

Appendix 1 Survey protocol

1. A. Socioeconomic variables used for defining the SEC

Categorical variables are indicated as either binary (B), ordinal (O) or nominal (N). Table from Hemerijckx et al., (2020, p.7).

Variable Collection	Numeric Variables	Categorical Variables
Household characteristics	Total number of household members Number of children (< 18 y.o.) Number of adult women (≥ 18 y.o.) Average commuting time Average education level Number of years lived in Kampala	Household tribe (N) Most spoken language (N) Urban agricultural activity (B) Housing type (N) Roofing type (N) Toilet type (N) Road type in front of home (N) Water source (13 dummy var.) (B) Energy source (9 dummy var.) (B) Cooking energy source (7 dummy var.) (B)
Neighborhood characteristics	Distance to nearest water source	Parish name (N) Neighborhood reputation (O) Neighborhood cleanliness (O) Neighborhood safety (O) Gated home infrastructure (O) Tarmacked road infrastructure (O) Flooding prevalence (O) Overall happiness in neighborhood (O)
Income and ownership	Income (2 var.) Workers employed at household Food expenditure (2 var.) Vehicle ownership (5 var.)	Tenure status (N) Ownership of air-conditioning (B) Ownership of a radio (B) Ownership of a television (B) Online activity (3 var.) (B) Ownership of a telephone (3 var.) (B)

1. B. Food Insecurity Experience Scale questionnaire

FAO questionnaire (Ballard et al., 2014, p.39), English version. Answers are yes/no.

During the last 12 months, was there a time when:

Question number	Variable	Survey Question
1	WORRIED	You were worried you would not have enough food to eat because of a lack of money or other resources?
2	HEALTHY	You were unable to eat healthy and nutritious food because of a lack of money or other resources?
3	FEWFOOD	You ate only a few kinds of foods because of a lack of money or other resources?
4	SKIPPED	You had to skip a meal because there was not enough money or other resources to get food?
5	ATELESS	You ate less than you thought you should because of a lack of money or other resources?
6	RUNOUT	Your household ran out of food because of a lack of money or other resources?
7	HUNGRY	You were hungry but did not eat because there was not enough money or other resources for food?
8	WHLDAY	You went without eating for a whole day because of a lack of money or other resources?

1. C. Household Dietary Diversity Score questionnaire

FAO questionnaire (FAO, 2010, p.7), English version. Answers are yes/no for each food group.

Please describe the foods (meals and snacks) that you ate or drank yesterday during the day and night, whether at home or outside the home. Start with the first food or drink of the morning.

Question number	Food group	Examples
1	CEREALS	corn/maize, rice, wheat, sorghum, millet or any other grains or foods made from these (e.g. bread, noodles, porridge or other grain products) + insert local foods e.g. ugali, nshima, porridge or paste
2	WHITE ROOTS AND TUBERS	white potatoes (in any form), white yam, white cassava, Matooke, Irish Potatoes, or other foods made from roots
3	VITAMIN A RICH VEGETABLES AND TUBERS	pumpkin, carrot, squash, or sweet potato that are orange inside + other locally available vitamin A rich vegetables (e.g. red sweet pepper)
4	DARK GREEN LEAFY VEGETABLES	dark green leafy vegetables, including wild forms + locally available vitamin A rich leaves such as amaranth, cassava leaves, kale, spinach
5	OTHER VEGETABLES	other vegetables (e.g. tomato, onion, eggplant) + other locally available vegetables
6	VITAMIN A RICH FRUITS	ripe mango, cantaloupe, apricot (fresh or dried), ripe papaya, dried peach, watermelon, passion fruit and 100% fruit juice made from these + other locally available vitamin A rich fruits
7	OTHER FRUITS	other fruits, including wild fruits and 100% fruit juice made from these
8	ORGAN MEAT	liver, kidney, heart or other organ meats or blood-based foods
9	FLESH MEATS	beef, pork, lamb, goat, rabbit, game, chicken, duck, other birds, insects
10	EGGS	eggs from chicken, duck, guinea fowl or any other egg
11	FISH AND SEAFOOD	fresh or dried fish or shellfish
12	LEGUMES, NUTS AND SEEDS	dried beans, dried peas, lentils, nuts, seeds or foods made from these (eg. hummus, peanut butter)
13	MILK AND MILK PRODUCTS	milk, cheese, yogurt or other milk products
14	OILS AND FATS	oil, fats or butter added to food or used for cooking
15	SWEETS	sugar, honey, sweetened soda or sweetened juice drinks, sugary foods such as chocolates, candies, cookies and cakes
16	SPICES, CONDIMENTS, BEVERAGES	spices (black pepper, salt), condiments (soy sauce, hot sauce), coffee, tea, alcoholic beverages

1. D. Food accessibility

Variable	Survey Question
Food expenditure	What is your household’s average daily expenditure on food?
Household income	What is the total monthly household income in UGX?
Proximity to most used food source	How long does it take you to travel to [most frequented market by respondent]?
Urban agricultural activity	Does this household produce any kind of food or other agricultural products, whether sold or consumed?
Food sharing activity	Do you receive or give food from/to neighbours, family members or other sources?
Tribe	Which tribe (or if not Ugandan: nationality) do the members of your household identify as?

Appendix 2 Sampling information (Hemerijckx et al., 2020)

Between July and December 2019, six surveyors approached homes in 15 contrasting parishes (the small administrative units, SAU) of the Greater Kampala Metropolitan Area (GKMA). We aimed to survey households at SAU that are contrasting both in terms of their location within the GKMA and in terms of socioeconomic population dynamics. The socioeconomic information was based on prior geographic research on the population of the SAU by Vermeiren et al. (2016), combined with consulting local experts at the Urban Action Lab of Makerere University. In addition, to select the final sampled parishes, we considered neighbourhood accessibility and surveyor safety. The sample size was thus limited by the practical access to the SAU. Within the selected SAU, a snowball strategy was adopted to sample households. This implied that a local council representative, after giving their informed consent, led the surveyors to households and assisted surveyors to clarify the purpose of the research.

As such, a convenience sampling method was adopted on the field. However, as a guideline to target sample size we calculated that with a desired confidence interval of 95%, the sample size of 541 households results in a margin of error of 1.97% based on Cochran's (1963) sample size formula for estimating prevalence:

$${n}_{0}= \frac{{Z}^{2}p(1-p)}{{e}^{2}}$$

(2)

where:

• n is the sample size (541 households, with 2487 individuals).

• p is the population proportion (assumed at 0.5 for complete uncertainty).

• Z the Z-score (1.96 for a confidence interval of 95%).

• e is the error margin (1.97%).

The final sample contained 541 households. Information was gathered on a total of 2487 individuals within these households. Prior to clustering the households by socioeconomic group, 16 households that exceeded a threshold of 37% of missing data (i.e. the respondent did not respond to over 37% of the survey questions) were excluded from analysis. Participants had an average of 4.25% missing data (SD = 10.91%, range = 0–95%). This threshold minimises the number of households that would be excluded, while also balancing adequate sampling per participant. This way, 525 households were included in the final dataset that was used for the present study. Figure 6 below shows the spatial distribution of the sampled households compared to the formal food markets as registered on OpenStreetMap.

Fig. 6

Spatial distribution of the input data

Appendix 3 Comparison of the food group categories used in the FAO dietary diversity questionnaire (16 food groups), the required aggregation to obtain the Household Dietary Diversity Score (HDDS, 12 food groups) (FAO, 2010), and the Uganda food consumption census data from the national household survey 2016/2017 (9 food groups) (UBOS, 2018)

FAO dietary diversity questionnaire	HDDS aggregation	Uganda national household survey
Cereals	Cereals	Staples (cereals, roots and tubers)
White roots and tubers	White roots and tubers
Vitamin A rich vegetables and tubers	Vegetables	Vegetables
Dark green leafy vegetables
Other vegetables
Vitamin A rich fruits	Fruits	Fruits
Other fruits
Organ meat	Meat	Meat, fish, eggs
Flesh meats
Eggs	Eggs
Fish and seafood	Fish and seafood
Legumes, nuts and seeds	Legumes, nuts and seeds	Pulses and nuts
Milk and milk products	Milk and milk products	Milk
Oils and fats	Oils and fats	Oil and fats
Sweets	Sweets	Sugar
Spices, condiments, beverages	Spices, condiments, beverages	Spices

Appendix 4 Results of the Rasch model

Item	Severity				Standard Error				Infit
SEC	EH	EL	NM	NL	EH	EL	NM	NL	EH	EL	NM	NL
WORRIED	-2.32	-1.75	-1.55	-1.61	0.31	0.41	0.38	0.30	0.99	1.10	1.14	1.12
HEALTHY	-0.78	-0.72	-1.10	-0.20	0.31	0.38	0.36	0.26	1.08	1.68	1.14	1.51
FEWFOOD	-1.70	-1.35	-1.55	-1.11	0.31	0.40	0.38	0.28	0.99	0.75	0.82	0.83
SKIPPED	0.51	-0.36	0.32	0.13	0.35	0.37	0.35	0.25	0.82	0.79	0.97	0.83
ATELESS	-0.36	-0.84	-0.48	-0.49	0.32	0.38	0.34	0.26	1.06	0.83	0.89	0.87
RUNOUT	1.08	0.81	0.94	0.55	0.38	0.38	0.36	0.25	1.26	0.62	0.71	0.82
HUNGRY	1.08	1.40	0.62	0.97	0.38	0.39	0.35	0.25	0.86	0.83	1.05	0.83
WHLDAY	2.50	2.82	2.78	1.76	0.48	0.51	0.50	0.27	1.12	1.25	1.19	1.25

SEC	Rasch reliability
EH	0.748
EL	0.751
NM	0.732
NL	0.715

1. All infit statistics within the range of 0.7 – 1.3 imply that the scale is adequately assuming equal discrimination between items. Although 3 infit statistics exceed this range (indicated in bold) overall the scale performed well. The Rasch reliability, which describes the proportion of the total variance accounted by the model, exceed the minimum acceptable value of 0.7 for each SEC.