1
Takefumi Fujimoto (presenter)
Aya Suzuki
Department of International Studies
Graduate School of Frontier Sciences
The University of Tokyo
Different Strategies of Crop Diversification
Between Poor and Non-Poor Farmers
-Concepts and Evidence from Tanzania-
(Two agricultural plots )
Research topic: crop diversification (CD) in Tanzania
→ Farmers grow diverse variety of crops, rather than one crop
Ex) maize, rice, sorghum, beans, tomato, carrot, cotton …
2
MAIZE ONLY
Farmers grow only maize
→ No crop diversification
MAIZE
BEANS
TOMATO
Sorghum
Farmers grow diverse crops
→ Crop diversification
2
Two benefits of CD:
(1) To reduce agricultural risks
・Wide range of planted crops provides wide adaptability to agronomic shocks (e.g., drought, crop diseases)
・Diverse composition of harvested crops provides robustness to price fluctuation in markets
(2) To increase income among poor farmers
・Productivity increases by agronomic synergies between crops (e.g., intercropping legumes)
→ CD has been empirically found to benefit the poorest the most (Michler & Josephson 2017; Asfaw et al. 2018; Asfaw et al. 2019; Tesfaye & Tirivayi 2020)
3
(Purpose)
To clarify poor and non-poor farmers adopt different strategies of CD
・Poor farmers pursue CD for robust food securities
・Non-poor farmers adopt CD to stabilize market income
This presentation:
(1) Explains why poor & non-poor farmers have different motivations for CD
(2) Empirically shows poor & non-poor farmers adopt CD differently:
・poor farmers respond to shocks on food securities
・Non-poor farmers respond to shocks on market income
4
Contents
5
Conceptual Framework
→ farmers’ utility maximization framework
We must consider two issues:
(1) How to express difference between poor & non-poor farmers
→ Poor farmers with low income prioritize consuming particular goods for survival
(2) How to express risk-reducing nature of CD
→ Crop diversity in farmland alleviates uncertainty
6
7
8
9
Budget line
Indifference curve
→ Solution in customary consumer problem
10
Subsistence line
Must be consumed
above subsistence line
11
12
13
Conceptual Framework
Hypothesis:
Poor/non-poor farmers adopt CD in a different manner
Three potential factors behind different motivations for CD:
・Subsistence requirement (poverty traps v.s. risk dispersion)
・Balanced dietary intakes
・Stabilization of market income
14
Subsistence requirement
Priority of poor farmers:
→ To satisfy the subsistence constraint in any state
Poor farmers may or may not adopt CD
Poor farmers specialize in producing high-calorie crops for survival
(2) Risk dispersion (CD↑)
Poor farmers adopt CD to disperse risks
15
Poverty traps (CD↓)
16
Poverty traps (CD↓)
17
Poverty traps (CD↓)
18
Risk dispersion (CD↑)
19
Risk dispersion (CD↑)
20
Subsistence requirement
Poor farmers may or may not adopt CD for survival
(1) Poverty traps
High crop diversity involves risks of insufficient energy intakes
⇒ CD↓
(2) Risk dispersion
Specialization in a certain crop involves risks of heavy damages
⇒ CD↑
21
Balanced dietary intakes
22
Balanced dietary intakes
23
Balanced dietary intakes
Moreover, income increase frees consumption restriction by the subsistence constraint
→ Poor farmers can choose consumption as they prefer
24
Stabilization of market income
The subsistence constraint is NOT binding among non-poor farmers
→ Non-poor farmers can consume luxury food and non-food items
25
⇒ Non-poor farmers try to stabilize market income used for their purchases by CD
Summary of Conceptual Framework
Three potential factors behind CD between poor/non-poor farmers
・+ : CD increases
・- : CD decreases
Farmers experienced shocks in the past
→ They recognize possible risks in relation to three factors
→ They determine crop diversity to address future risks
26
| Subsistence Requirement | Balanced Dietary Intakes | Stabilization of Market Income | |
| Poverty Traps | Risk Dispersion | | |
Poor Farmers | - | + | + | |
Non-Poor Farmers | | | | + |
Contents
27
Dataset
28
Tanzania National Panel Survey (LSMS-ISA)
・Nationwide questionnaire for households
・Collect socioeconomic & agronomic information
Farmers in Tanzania’s mainland
→ Balanced panel of 1,332 farmers in each 2010 & 2012
Indicators of Crop Diversification
29
Indicators of Crop Diversification
30
Methodology
31
Methodology
32
33
Contents
34
35
Type of CD indicator | Estimated Threshold (shillings) |
Simple Count of Crop Number | 206,276 |
Simpson Index | 204,938 |
Results (Simple Count of Crop Number)
Poor farmers increase crop number against past drought/flood
→ Poor farmers may build farming systems robust to heavy damages
⇒ Subsistence requirement (risk dispersion)
36
Experience of agricultural shock | Below threshold (poor farmer) | Above threshold (Non-poor farmer) |
Drought/flood | 0.28* (0.15) | 0.02 (0.07) |
Crop disease/pests | -0.21 (0.17) | 0.10 (0.08) |
Large fall in sales prices for crops | -0.06 (0.14) | 0.13* (0.08) |
Large rise in price of food | 0.25* (0.14) | 0.06 (0.06) |
Large rise in input price | -0.04 (0.15) | 0.20** (0.08) |
Livestock loss (Died or stolen) | -0.17 (0.21) | 0.07 (0.10) |
(Standard errors in parentheses)
Results (Simple Count of Crop Number)
Poor farmers increase crop numbers against rise in food price
・Food price↑⇒consumption↓
→ Poor farmers address a threat to food security
⇒ Subsistence requirement (risk dispersion)
& Balanced dietary intakes
37
Experience of agricultural shock | Below threshold (poor farmer) | Above threshold (Non-poor farmer) |
Drought/flood | 0.28* (0.15) | 0.02 (0.07) |
Crop disease/pests | -0.21 (0.17) | 0.10 (0.08) |
Large fall in sales prices for crops | -0.06 (0.14) | 0.13* (0.08) |
Large rise in price of food | 0.25* (0.14) | 0.06 (0.06) |
Large rise in input price | -0.04 (0.15) | 0.20** (0.08) |
Livestock loss (Died or stolen) | -0.17 (0.21) | 0.07 (0.10) |
(Standard errors in parentheses)
Results (Simple Count of Crop Number)
Non-poor farmers increase crop number against past fall in crop price & rise in input price
・Crop price↓ ⇒ Income ↓
・Input price↑ ⇒ Cost ↑
→ Non-poor farmers address threats to market income
⇒ Stabilization of market income
38
Experience of agricultural shock | Below threshold (poor farmer) | Above threshold (Non-poor farmer) |
Drought/flood | 0.28* (0.15) | 0.02 (0.07) |
Crop disease/pests | -0.21 (0.17) | 0.10 (0.08) |
Large fall in sales prices for crops | -0.06 (0.14) | 0.13* (0.08) |
Large rise in price of food | 0.25* (0.14) | 0.06 (0.06) |
Large rise in input price | -0.04 (0.15) | 0.20** (0.08) |
Livestock loss (Died or stolen) | -0.17 (0.21) | 0.07 (0.10) |
(Standard errors in parentheses)
Results (Simpson Index)
Poor farmers increase SI against drought/flood &
rise in food price
→ Consistent with results of crop numbers
・build robust farming systems
・address a threat to food security
39
Experience of agricultural shock | Below threshold (poor farmer) | Above threshold (Non-poor farmer) |
Drought/flood | 0.38*** (0.15) | -0.00 (0.07) |
Crop disease/pests | -0.20 (0.18) | 0.07 (0.08) |
Large fall in sales prices for crops | 0.05 (0.17) | 0.04 (0.08) |
Large rise in price of food | 0.37** (0.16) | 0.03 (0.06) |
Large rise in input price | -0.17 (0.16) | 0.09 (0.07) |
Livestock loss (Died or stolen) | -0.35* (0.20) | 0.04 (0.09) |
(Standard errors in parentheses)
Results (Simpson Index)
Poor farmers decrease SI against livestock loss
・livestock is labor for tillage & seed planting
→ Poor farmers cannot repurchase livestock
→ low ability of cultivation
→ CD is discouraged (Subsistence requirement: poverty trap)
40
Experience of agricultural shock | Below threshold (poor farmer) | Above threshold (Non-poor farmer) |
Drought/flood | 0.38*** (0.15) | -0.00 (0.07) |
Crop disease/pests | -0.20 (0.18) | 0.07 (0.08) |
Large fall in sales prices for crops | 0.05 (0.17) | 0.04 (0.08) |
Large rise in price of food | 0.37** (0.16) | 0.03 (0.06) |
Large rise in input price | -0.17 (0.16) | 0.09 (0.07) |
Livestock loss (Died or stolen) | -0.35* (0.20) | 0.04 (0.09) |
(Standard errors in parentheses)
Summary of results
Poor and non-poor farmers have different motivations for CD
Poor farmers:
・Poor farmers adopt CD for food security
・Poor farmers may be trapped in poverty due to loss of productive assets
Non-poor farmers:
・Non-poor farmers adopt CD to stabilize income obtained from market transactions
41