Food security vulnerability in South Africa Case study Limpopo
Content I. II. III. IV. V.
Project introduction Methodology General results Food security determinants Policy priorities
Content I. II. III. IV. V.
Project introduction Methodology General results Food security determinants Policy priorities
Different actors & partners
Objectives • Identify the factors influencing food security (vulnerability) at household level and at municipality level • Compute a Food Security Index (FSI) based on four major components:
AVAILABILITY
ACCESSIBILITY
UTILISATION
STABILITY
Objectives lead to… • Development of an accessible assessment tool to measure food security vulnerability • Policy advice and new strategies related to food security
Content I. II. III. IV. V.
Project introduction Methodology General results Food security determinants Policy priorities
Methodology • Data source and collection: – Secondary data – Surveys (field data collection) • Survey on food security & determinants • Survey background information, municipality level
Qualitative and quantitative data
Overview questionnaire 1. Survey identification 2. Household demographics 3. Food availability & consumption 4. Agricultural production 5. Household income & expenditure 6. Characteristics of household 7. Stresses, shocks & coping strategies
Limpopo N
Musina Mutale
Vhembe Thulamela
Makhado
Blouberg
N1 Greater Giyani Greater Letaba Aganang Mogalakwena
Lephalale
Molemole
Capricorn
Waterberg
Mopani Greater Tzaneen
Ba-Phalaborwa
Polokwane Maruleng
Thabazimbi
Fetakgomo Ñ
Legend Road N1 Municipality Boundaries District Boundaries
Lepelle-Nkumpi
Modimolle
Mookgopong
Bela-Bela
Makhudutamaga
Bohlabela
Tubatse Bushbuckridge
Sekhukhune
Marble Hall
Groblersdal
Communities District Rates 15% - 20% 20%
25
0
25
50
75 Kilometers
Scale 1 : 2 500 000
Data collection (28/07- 13/08) DISTRICT
MUNICIPALITY 1
MUNICIPALITY 2
Capricorn
Blouberg (60)
Molemole (60)
Total: 120 surveys Mopani Total: 120 surveys Sekhukhune Total: 120 surveys Vhembe Total: 120 surveys Waterberg Total: 120 surveys
data collection data entry Giyani (60)
4 enumerators + 1 student 1 student
Tubatse (60)
data collection data entry Mutale (60)
4 enumerators + 1 student 1 student
Maruleng (60)
data collection data entry Fetakgomo (60)
RESPONSIBLE?
4 enumerators + 1 student 1 student
Thulamela (60)
data collection data entry
4 enumerators + 1 student 1 student
Mogalakwena (60) Mookgopong (60) data collection data entry
4 enumerators + 1 student 1 student
post- data collection phase… • 15th of August – 25th of August: – Data analysis • Obtained results? • Interpretation of results?
– Writing of report
Content I. II. III. IV. V.
Project introduction Methodology General results Food security determinants Policy priorities
General results • • • • • • •
Food security & Poverty Human capital Food production Access to resources Household income Food consumption pattern Shocks & stresses
Food security & poverty in Limpopo • 53% severely food insecure • 32% less then 1$ per day • 60% less then 2$ per day
21%
Food secure
53% 26%
Moderately food insecure Severely food insecure
Validating Food Security Study results… • In 2002 it was suggested that 43% of SA households suffer from food poverty A state where physiological human needs are not adequately met as the available amount of money is not enough to purchase a basic nutritionally balanced diet
National Food Consumption Survey -Fortification Baseline Limpopo, 2011 (NFCS-FB), 2005 1 in 2 hh experienced hunger 53% severely food insecure 1 in 3 hh were at risk of hunger 26% moderately food insecure 1 in 5 ppl were food secure 21% food secure
National Values vs. Food Security Study South African situation: • 50% of the 10 to 11 million households in South Africa can be classified as low-income households
Prahalat & Hart, 2006 / NFCS, 2005
Household size
US$ / Per capita / per day
5
$0.59
StatsSA, 2005/2006
$0.60 (Poorest 10%) $1.32 (2nd poorest 10%)
Current study
32% < $1 60% < $2
6-7
Food security status district level waterberg district
18.8%
vhembe district
21.2%
15.4%
65.8%
33.1%
45.8%
Food secure
sekhukhune district
36.8%
34.2%
29.1%
Moderately food insecure Severely food insecure
mopani district
12.4%
capricorn district
14.4%
0%
24.0%
63.6%
24.6%
20%
61.0%
40%
60%
80%
100%
• Highest food insecurity levels in Waterberg & Mopani district • Lowest food insecurity levels in Sekhukhune district
Food security municipality level 25.4%
Mogalakwena
12.1%
Mookgopong
25.9%
10.0% 14.8%
Giyani
10.6%
0%
31.0%
moderately food insecure
25.9%
severely food insecure
65.0%
23.0%
62.3%
21.2%
59.6%
27.3%
20%
food secure
32.2%
25.0%
19.2%
Molemole
41.7% 37.3%
43.1%
Fetakgomo
50.0%
40.0%
30.5%
Tubatse
Blouberg
67.2%
18.3%
Mutale
64.4%
20.7%
24.1%
Thulamela
Maruleng
10.2%
62.1%
40%
60%
80%
100%
• Highest food insecurity levels in Mookgopong, Maruleng & Mogalakwena • Lowest food insecurity levels in Fetakgomo & Tubatse
Number of hungry months 60
50
Percentage
40
30 Percent
20
10
0 0
1
2
3
4
5
6
7
Average number of hungry months
8
9
10
11
12
Poverty per district
Districts of Limpopo province
Waterberg
Vhembe
Percent of people living on less than 2 $US per day
Sekhukhune
Percent of people living on less than 1 $US per day Mopani
Capricorn
0
10
20
30
40 Percentage
50
60
70
80
Poverty on district level waterberg district
69%
31%
vhembe district
81%
19%
sekhukhune district
79%
21%
> 1 US$ per day < 1 US$ per day
mopani district
50%
capricorn district
50%
62%
0%
20%
38%
40%
60%
80%
100%
• Highest poverty rates in Mopani District • Lowest poverty rates in Vhembe & Sekhukhune District
Poverty on municipality level 34%
66%
Mogalakwena
28%
72%
Mookgopong
25%
75%
Thulamela
13%
87%
Mutale
17%
83%
Tubatse
25%
75%
Fetakgomo 43%
Maruleng
43%
Molemole
63%
37%
Blouberg
62%
38%
0%
20%
1 US$ per day
40%
60%
• Highest poverty rates in Maruleng & Giyani
80%
100%
Human capital Human capital indicators
General
Household size
6-7 (3)
Education level (share of hhold head with no schooling)
33%
Gender (share of female headed hholds)
40%
Dependency ratio (income earners/ total hhsize)
0.85 (0.18)
Migrant workers (share of hholds with contributing migrant)
25,5%
Education level (hhold head) No schooling waterberg
33%
Junior primary vhembe
13%
sekhukhune
Senior primary Some Secondary
36%
mopani
Completed high school
42%
capricorn
Courses or certificates for formal training Diploma or degree
39%
0%
20%
40%
60%
80%
100%
• Overall education levels are lowest in Mopani and highest in Vhembe
Education level 35
30
25
Percentage
20
15
10
5
0 No schooling
Junior primary
Senior primary
Some Secondary
Completed high school
Courses or certificates
Diploma or degree
Activity rate 0.2 0.18 0.16
Ratio of income earners
0.14 0.12 0.1
Mean
0.08 0.06 0.04 0.02 0 vhembe
waterberg
capricorn district
sekhukhune
mopani district
Food production • 57% of households involved in crop production • 50% of households involved in livestock production • Most popular crops: – – – – –
Maize & Mango: 27% Pawpaw:15% Spinach: 15% Tomatoes & Oranges: 13% Banana & Guava: 10%
• Most popular animals: – Poultry: 50% – Cattle & goats: 22%
Crop production Share of total households (N=599) 80% 70% 60%
capricorn
50%
mopani
40%
sekhukhune vhembe
30%
waterberg
20%
general
10% 0% maize
mango
pawpaw
spinach
oranges
tomatoes
• Vhembe district: crop production most popular • Sekhukhune & Waterberg district: crop production less popular
Livestock production Share of total households (N=599) 70% 60% capricorn
50%
mopani
40%
sekhukhune
30%
vhembe
20%
waterberg
10%
general
0% poultry
goats
cattle
pigs
• Livestock production most popular in Vhembe district
Pork and offalalone Venison wild game Meat lamb goat and offal Red meat not part of a stew Mopani worms and insects Ham poloni cold meat tinned meat Beef and offal
Food Groups / Types
Legumes nuts & seeds Fish Roots & tubers Eggs Other fruits Vit. A fruits & vegetables Dairy products Other vegetables Poultry Other cereals Oil & butter Beverages Food products containing sugar Maize products 0
1
2
3
4
5
6
Mean frequency of consumption in the past seven days.
7
8
Venison wild game Pork and offalalone Mopani worms and insects Meat lamb goat and offal Red meat not part of a stew Ham poloni cold meat tinned meat
Food groups/types
Legumes nuts & seeds Vit. A fruits & vegetables Other fruits Roots and tubers Other vegetables Oil & butter Beverages Eggs Fish Dairy Sugar Beef and offal Poultry Other cereals Maize products 0
50
100
150
Average Monthly Expenditure in Rands
200
250
Acces to resources Land resources
• Average land size 0.95 ha (SD: 2.36)
Water sources
• Yard tap: 33% • Public tap: 20% • Borehole: 20%
Energy Financial
• 92% is connected to electricity
• Burial insurance: 57% • Savings account: 42%
Household income • Average income per capita: 605 R/month (SD: 1200 R/month) • Vhembe highest, Mopani lowest income per capita 25 21.2
20.0 20
15.5 15 11.1 10
5
0
7.6
6.7 4.5
7.6 5.7
900 800 700 600 500 400 300 200 100 0
share of households (monthly income categories)
monthly income per capita
Monthly income distribution of households per district 100% 90% 80% 70%
R7500
farming income • Sekhukhune -> formal salary
Food consumption pattern Food expenditure pattern 2% 4%3% 4% 6% 6% 15% 15%
34%
cereals bread non red meat fruits & veggies red meat eggs roots & tubers dairy legumes
100% 80% 60%
10% 7%
•
Cereals, bread & non red meat -> 60% of food expenditure Red meat ->6% of food expenditure
4% 6%
16%
14%
19%
13%
17%
40% 20%
19% 40%
33% 21%
0% Food secure
•
5% 6%
•
Moderately food insecure
Severely food insecure
Food insecurity • Relatively more spend on cereals • Relatively less spend on meat &dairy
Shocks, stresses and coping strategies • Increase in food price is most important stress in the area • Most important coping strategies: – Borrowing money & food from relatives (social capital) – Reducing food consumption & spending – Only 7,5% looks for more employment opportunities
Intra household food distribution vulnerability (%) 50 45 40
Vulnerability (%)
35 30 25 Intra-household food distribution 20 15 10 5 0 Children
Older children
Female adults
Age Category
Male adults
Importance of coping strategies share of households that use coping strategy 0
10
20
30
40 40.8
borrow money from relatives/friends 33.3
borrow food from relatives/friends
32
reduce food consumption
31.5
reduce spending 18.9
selling livestock 14.9
use own savings
13.7
receive grants or gifts look for additional employment take out loan from mashionisa take out loan from formal institution
7.5 6.9 5.7
50
Content I. Project introduction II. Methodology III. General results IV. Food security determinants V. Policy priorities
Food security determinants • Description of different food security categories • What is the difference between food secure and food insecure households?
• Who has highest probability of being food insecure?
• What are the determinants of food insecurity?
Determinants of food security Human capital
Farming system
Food security Household income
Access to resources
Overview determinants Human capital • Household size • Education level • Gender head • Dependency ratio • Migrant workers
Farming system • Subsistence food production • Livestock production
Access to resources • Land • Water • Schooling
Household income • Income per capita • Remittances per capita • Type of income
Identification determinants • Two different multivariate analysis – Regression analysis – Cluster analysis
• Different methods – Check for robustness of findings
Outcome regression analysis Coëfficient constant
Test value 5.42***
HUMAN CAPITAL Household size
0,202
5,21***
Age household head
-0.19
-4.37***
Education level (household head)
-0.23
-4.69***
Gender (household head)
0.08
2.06**
Dependency ratio (income earners/total hhsize)
0.004
0.092
Maizeproduction (dummy)
-0.01
-0.27
Mango production (dummy)
-0.05
-1.14
Pawpaw production (dummy)
-0.04
-1.03
Spinach production (dummy)
-0.08
-1.85*
Tomatoe production(dummy)
-0.01
-0.22
Cattle (dummy)
-0.06
-1.55
Goats (dummy)
0.04
0.91
Poultry (dummy)
0.06
1.45
FOOD PRODUCTION
Outcome regression analysis Coëfficient
Test value
Cropping land size (ha)
-0.05
-1.24
Distance to water source (m)
0.08
2.24**
Monthly income per capita (Rand/month)
-0.09
-2.14**
Formal income (dummy)
-0.12
-2.51**
Grants & gifts (dummy)
0.10
2.12**
Unskilled labour income (dummy)
0.17
4.65***
Remittances (dummy)
-0.16
-4.19***
Skilled labour or entrepreneurial activity (dummy)
0.05
1.18
Farm income(dummy)
-0.07
-1.60
ACCESS TO RESOURCES
HOUSEHOLD INCOME
*10% significance level, **5% significance level, ***1% significance level.
• • •
Independent HFIAS score (Food insecurity score) Negative coëfficients result in higher food security levels R= 0.57; R2=0.32
Cluster analysis – Creating clusters using different variables – Looking for overlap between different determinants & different indicators of food security – Different types of variables included
Human capital
Food production
Access to land
Food security indicators
Food security indicators •Food insecurity score •Importance of food in total expenditure (%) •Importance of cereals in food expenditure (%)
Different clusters Cluster 1 (N=384) 2 7.4 0.87 0.8 2.1 2.1 240 0.86 12 0.62
Education level (1-7) Total household size Dependency ratio Land size (ha) Crop index (Σ crops cultivated) Livestock index (Σ different animal types) Income per capita (Rand/month) Grants & gifts as income source (dummy) Food insecurity score (0-27) Food expenditure (share of total monthly expenditure) Expenditure on cereal (share of total monthly food 0.39 expenditure) *10% significance level, **5% significance level, ***1% significance level.
Cluster 2 (N=132) 3 5 0.81 1.3 2.4 2.8 830 0.64 6 0.54
Cluster 3 (N=25) 5 5.1 0.67 1.5 3.6 5.0 1900 0.32 4 0.37
0.24
0.17
Test 38.56*** 11.82*** 34.14*** 0.95 2.34* 25.13*** 1898*** 17.66*** 27.38*** 15.35***
Different clusters Education level (1-) Total household size Dependency ratio Land size (ha) Crop index (Σ crops cultivated) Livestock index (Σ different animal types)
Cluster 1 (N=384) 2 7.4 0.87 0.8 2.1 2.1
Cluster 2 (N=132) 3 5 0.81 1.3 2.4 2.8
Cluster 3 (N=25) 5 5.1 0.67 1.5 3.6 5.0
Income per capita (Rand/month) Grants & gifts as income source (dummy)
240 0.86
830 0.64
1900 0.32
Food insecurity score (0-27) Food expenditure (share of total expenditure)
12 0.62
6 0.54
4 0.37
Expenditure on cereal (share of total food expenditure) 0.39
0.24
0.17
Low education level, Least land and low crop & livestock indices Lowest income, most dependent on grants & gifts High food insecurity score, High importance of food & staple foods in expenditure
Different clusters Education level (1-) Total household size Dependency ratio Land size (ha) Crop index (Σ crops cultivated) Livestock index (Σ different animal types)
Cluster 1 (N=384) 2 7.4 0.87 0.8 2.1 2.1
Cluster 2 (N=132) 3 5 0.81 1.3 2.4 2.8
Cluster 3 (N=25) 5 5.1 0.67 1.5 3.6 5.0
Income per capita (Rand/month) Grants & gifts as income source (dummy)
240 0.86
830 0.64
1900 0.32
Food insecurity score (0-27) Food expenditure (share of total expenditure)
12 0.62
6 0.54
4 0.37
Expenditure on cereal (share of total food expenditure) 0.39
0.24
0.17
Higher education level, More land and higher crop & livestock indices Higherincome, less dependent on grants & gifts Lowest food insecurity score, Lower importance of food & staple foods in expenditure
Different clusters Education level (1-) Total household size Dependency ratio Land size (ha) Crop index (Σ crops cultivated) Livestock index (Σ different animal types)
Cluster 1 (N=384) 2 7.4 0.87 0.8 2.1 2.1
Cluster 2 (N=132) 3 5 0.81 1.3 2.4 2.8
Cluster 3 (N=25) 5 5.1 0.67 1.5 3.6 5.0
Income per capita (Rand/month) Grants & gifts as income source (dummy)
240 0.86
830 0.64
1900 0.32
Food insecurity score (0-27) Food expenditure (share of total expenditure)
12 0.62
6 0.54
4 0.37
Expenditure on cereal (share of total food expenditure) 0.39
0.24
0.17
Higher education level, More land and higher crop & livestock indices Medium income, still very dependent on grants & gifts Lower food insecurity score, Still high importance of food in expenditure
Clusters vs HFIA category • Distribution of HFIA categories in different clusters 100% 90%
31%
24%
Share of households
80% 70%
54%
63%
60%
24% severely food insecure
27%
50%
moderately food insecure
40%
food secure
30%
26% 26%
20% 10%
42%
52% 20%
10%
0% cluster 1 Χ2 : 86,51***
cluster 2
cluster 3
general
Clusters vs HFIA category Food secure
Moderately food insecure
Severely food insecure
Total
Cluster 1
10
27
63
100
Cluster 2
42
27
31
100
Cluster 3
24
24
52
100
General
20
26
54
Identification determinants • From these analysis we find the most important determinants of food security in Limpopo area Household Income
Education level
Dependency on grants & gifts
Household food security status
Type of employment
Content I. II. III. IV. V.
Project introduction Methodology General results Food security determinants Policy priorities
Most important policy priorities • Based on determinants of food security certain policy priorities can be distinghuised • Not all determinants can be tackled directly through policy
Policy priorities Low education level High dependency ratio
Vulnerable female headed household Low household income High dependency on grants & gifts
Food production
• Promote education in rural areas
• Decrease by ensuring job opportunities & facilitating the labour market • Support female headed households • Special focus on girls and women in rural development policies • Promote employment • Facilitate labour market • Ensure sustainability of income • Promote employment • Manipulate incentives • Modify grant system • Promote the potential for household food production to contribute to food security