The Relationship between Collective Action and Intensification of Livestock Production: The Case of northeastern Burkina Faso

1
The Relationship between Collective Action and
Intensification of Livestock Production The
Case of northeastern Burkina Faso 

• Nancy McCarthy and Celine Dutilly-Diane
•  
• Research Fellow, International Food Policy
  Research Institute, Washington, DC
• Researcher, CIRAD, Montepellier, France
•  

2
Presentation Outline

• Background Burkina Faso
• Theoretical Model of Intensification
• Social Optimum vs. Non-Cooperative Game
• Incorporating Cooperation
• Empirical Results
• Households use of purchased livestock inputs
• Vaccines, agro-industrial by products, crop
  residues
• Concluding Observations

3

Mali
Study Region
Niger
Vegetation Steppe, acacia dominant Rainfall
Long-term mean 400-500mm monomodal
variability CoV .2-.3
4
Context Agriculture Production Systems in
Northeastern Burkina Faso
Crop Livestock both important 46
Community land in (common) pastures (44 of
48 also had access to local pastures shared
with neighboring villages) 34 in Crops
20 in dunes, denuded land, ponds,
homesteads Mobility also important More
than ½ of communities had some herds mobile
during some part of year Though, in 1999-00,
not at home 15 of time Cash Incomes 47
Livestock products, 27 remittances, 6 crops,
6 salary, 13 other (petty trade, handicrafts)
5
Theoretical Model of Input Use Social Optimum
vs. Non-Cooperation
Non-Cooperation
Social Optimum
6
Theoretical Model of Input Use Social Optimum
vs. Non-Cooperation
Non-Cooperation
Social Optimum

At the same
7
Four Cases
1. If
2. If
3. If
4. If
8
Hypotheses from Model

• IF overstocking on common pastures, and this
  drives down marginal returns to purchased inputs,
  then their use also declines (health inputs,
  high-value agro-industrial by-products)
• IF overstocking on common pastures, and this
  drives up marginal returns to purchased inputs,
  then their use increases (crop residues)
• IF? no cooperation

9
On to Cooperative Capacity Community-Level
Indicators
Scoring Coefficients Factor Analysis
 
10
Cooperative Capacity Explanatory Factors
Table A1 Scoring coefficients for first two
factors of community level cooperative capacity
indicators

• Community-Level Factors
• Households, Households2 (,-) Total land
  endowment (?)
• Wealth and Ethnic Heterogeneity
• Migration for Wage Work
• Public Education
• External Projects
• Before 1993 (technical, input/commodity specific)
• After 1993 (village-level systems
  approach,institutions)
• Structure of NRM institutions
• Dominance of Chief
• Collaborative Rule Making
• External Pressure on Home Resources (neighbors,
  transhumants)
• Proximity to regional capital

 
11
Cooperative Capacity Explanatory Factors
Table A1 Scoring coefficients for first two
factors of community level cooperative capacity
indicators

• Network Capacity Greater in moderately
  well-populuted, heterogeneous communities located
  in the variable and low rainfall areas relatively
  close to the regional capital. Also greater
  where collaboration between chief and household
  members, and the number of programs which begun
  in the earliest period (before 1993).
• Active Capacity Greater in more favorable and
  less variable environments, where wealth more
  equally distributed, less migration for wage
  work, and with more external projects begun after
  1993.
• Overgrazing Calculated from community-level
  stock density equations, setting Active Capacity
  1.

 
12
Probits Purchase of Vaccines
13
Probits Purchase of Agro-Industrial By-Products
14
Probits Purchase of Millet and Sorghum Residue
15
Concluding Observations

• If we think of intensifying livestock production
  as increasing purchased goods per animal unit,
  theoretically, intensification will be lower
  where pastures are not well-managed instead
  substitute inputs for forage will increase
• Econometric Results support the hypothesis that
  intensification is lower where common pastures
  are not well-managed
• Intensification is lower in riskier environments
  (low mean and high variability of rainfall)
• Wealthier households (more assets) who own a
  greater proportion of herd managed, more likely
  to substitute for pasture AND purchase
  complimentary inputs

16
In Context of Other Results

• Higher active capacity, lower stock densities,
  more mobility, less land allocated to crops.
  Higher network capacity, higher stock densities,
  more land allocated to crops
• High active capacity, greater livestock incomes
  (absolute and share of total incomes) and higher
  total incomes higher network capacity, higher
  crop income share, higher off-farm income shares,
  and no impact on total incomes (Dutilly-Diane,
  Sadoulet de Janvry
• Higher active capacity, or lower stock densities
  on common pastures, greater likelihood of
  purchasing complimentary inputs. Greater network
  capacity, or higher stock densities, more spent
  on low-value pasture substitutes