Transcript Modeling Smallholder Commercialization

AAMP Training Materials
Module 2.2: Modeling Smallholder Commercialization
Nicholas Minot (IFPRI)
[email protected]
Contents
• Agricultural commercialization conceptual framework
• Agricultural household models before and after 1985
– Separable vs Non-Separable household models
• Household modeling exercises
• Conclusions
• Resources
Agricultural Commercialization
Conceptual Framework
Food and cash
crop production
-
Food
demand
=
Marketed
surplus
Agricultural Commercialization
Conceptual Framework
Labor, mgt
Land
Weather
Inputs
Food and cash
crop production
Preferences
Population
-
Income
Food
demand
Food prices
Infrastructure and
market efficiency
International
markets & trade
=
Marketed
surplus
Agricultural household models – Before 1985
• “Separable household models”
– Maximize household income
– Allocate income among consumption goods to maximize utility
• “Separable” in that production decisions are not affected
by consumption decisions
• Marketed surplus is difference between optimal quantity
of production and optimal quantity of consumption
– They could only be equal by coincidence so this should only
occur rarely
Agricultural household models – Before 1985
Model doesn’t explain several important patterns
Prediction of separable model
Patterns in developing countries
Farmers specialize in few commodities
according to comparative advantage
Many farms are diversified, producing
many food and cash crops
Farmers allocate most land to
commercial crops
Small share of land allocated to cash
crop production
Farmers either buy or sell each crop
(only rarely does optimal production =
optimal consumption)
Many household grow and consume
staple crops, without buying or selling
Agricultural household models – Before 1985
• Attempts to explain differences between model and
actual farmer behavior
– Food first: Farmers first produce enough food for consumption
needs and only if they have extra land do they produce for the
market
• Problem: Overly simplified view since many farmers do produce
cash crops even though they have not produces all their food
requirements.
• 40-60% of smallholders in many countries are net buyers of staple
food crops
– Risk aversion: Farmers maximize income and minimize risk
• Problem: Better explanation, but still doesn’t explain the fact that
farmers do not participate in markets (neither buyer nor seller) for
many commodities
Agricultural household models – After 1985
• Non-separable household models
– Singh, Squire, & Strauss (1986) Agricultural Household Models
– Models incorporating transaction costs, ie cost of buying or
selling a commodity
– Creates a gap between a) price household can buy at and b)
price it can sell at
– Example:
• Suppose market price of maize is 300, but it costs 50 to transport it
to the market or to transport from market to household
• Farmer can pay 350 to purchase maize or he can earn 250 from
selling maize
• If maize is worth 290 to household, it is not worth buying or selling
maize, so household does not participate in market
Agricultural household models – After 1985
• Why do transaction costs make models non-separable?
– Cannot calculate profit-maximizing production without knowing
consumption preferences
– Consumption preferences  whether household is buyer or
seller  price of commodity for household  profit maximization
– Thus, consumption preferences and profit maximizing income
must be solve simultaneously
• Spatial arbitrage between household & market
– Shadow price (Ph) for household
•
•
•
•
Ph = Pm +t if household is buying commodity
Ph = Pm – t if household is selling commodity
Pm+t > Ph > Pm-t if household neither sells nor buys
The larger is t, the more likely household neither sells nor buys
Agricultural household models – After 1985
Implications of non-separable agricultural household models
Prediction of NON-separable model
Patterns in developing countries
If transaction costs are high, diversified
food consumption needs  diversified
food production patterns
Many farms are diversified, producing
many food and cash crops
If transaction costs high, farmers forced
to self-supply with food, meaning small
share of land allocated to cash crop
production
Small share of land allocated to cash
crop production
If transaction costs high, many
households do not participate in market
(Pm+t > Ph > Pm-t )
Many household neither buy nor sell
staple foods
Agricultural household models – After 1985
Implications of non-separable agricultural household models
Price
Price
Demand
Supply
Price range over
which household
does not react to
price changes
Supply
Demand
Quantity
One household
Quantity
Several households
Agricultural household models – After 1985
Implications of non-separable agricultural household models
Price
Price
Price range
where most
households
are sellers &
respond
Price range over which
many households are
self-sufficient & do not
respond
Price range over which
most households are
buyers & respond
Quantity
Several households
Quantity
Many households aggregated
Agricultural household models – After 1985
• Estimating non-separable household models
– Need data on production technology, market prices, transaction
costs, and consumption preferences
– Use simultaneous equations for supply and demand
– Econometric software such as Stata
• Simulation based on existing parameters
– Need info on current production and consumption patterns,
prices and transaction costs, and elasticities of supply and
demand
– Use mixed complementarity programming (MCP), which allows
inequalities (Ph <= Pm + t)
– Modeling software such as GAMS
– Very similar to modeling spatial equilibrium between markets
Agricultural household model in Excel
• Characteristics of Excel worksheet [Exercise – Ag HH Model]
–
–
–
–
–
Two commodities: maize and coffee
Area of maize & coffee depend on prices of both commodities
Yield of maize & coffee depend on fertilizer price
Maize demand depends on exogenous income
Can change: productivity (yield), world prices, exchange rate,
income, and distance (marketing costs)
• How to use the Excel model
–
–
–
–
–
BLUE represents cells you can change to calibrate model
YELLOW represents cells you can change to simulate a shock
GREEN shows the output cells, which should not be changed
Table shows “before” and “after” simulated shock
Solid lines represent “before”, dashed lines “after”
Agricultural household model – Exercises
1. Increase household income by 20% (Cell E23)
a)
b)
c)
d)
What happens to maize consumption?
Does the household start to buy maize? Why?
Why does coffee production decline?
What happens to marketed surplus?
2. Increase household income by 40%
a) What happens to maize consumption?
b) Does the household start to buy maize?
c) Explain why the household does or doesn’t start to buy maize
Agricultural household model – Exercises
3. Increase exchange rate by 20% (Cell E14)
– Increase the exchange rate from 85 KSh/US$ to 102 KSh/US$
(this implies a devaluation of the shilling devaluation)
a) Why does maize production decline?
b) Why does the fertilizer price increase?
c) What happens to coffee yield? And coffee area?
d) Explain the contradictory effects on coffee.
4. Increase exchange rate by 40% (Cell E14 = 119)
a) What happens to maize purchases?
b) Why does devaluation cause farmers to shift from maize to
coffee production?
5. Increase world fertilizer price by 20% (Cell E25 = 20%)
a) How is the effect different than 20% devaluation?
Agricultural household model – Exercises
6. Increase the marketing costs for fertilizer, maize, &
coffee by 30%
–
a)
b)
c)
d)
This simulates a household that lives farther from roads and markets
What happens to the cost of fertilizer?
What happens to the buying and selling price of maize?
What are the two factors causing coffee production to decline?
What happens to the marketed surplus percentage?
7. Increase the base maize consumption (Cell C7) from 1000 to
1400, then increase maize market cost by 50%
– This represents a net maize buyer and raises cost of purchasing maize
a) What happens to coffee production and sales?
b) Why do higher maize marketing costs reduce cash crop marketing
Agricultural household model – Exercises
8. Increase maize productivity by 30% (Cell C28)
– This simulates a 30% increase in maize productivity due to, for
example, an improved maize variety
a) What happens to maize yields?
b) What happens to maize area? Explain this change.
c) Why does coffee area and production increase?
d) What is the effect on the value of sales and the marketed
surplus?
9. Increase coffee productivity by 30% (Cell D28)
a) What happens to coffee yield and production?
b) What is the effect on the value of sales and the marketed
surplus?
c) Why does an increase in maize productivity have almost the
same effect on marketed surplus as an increase in coffee
productivity?
Agricultural household model – Exercises
10. Increase maize market price by 30% (Cell C11 = 36.4)
– This simulates a 30% increase in maize prices in the market due
to, for example, a bad harvest in another part of the country
a) What happens to maize area and yields?
b) What happens to coffee production?
11. Decrease maize market price by 10%
a) What is the effect on maize and coffee production?
b) Why do these price changes have so little effect on the
agricultural household?
12. Increase world coffee price by 20%
a) What happens to coffee yield and production? Why?
b) What is the effect on the value of sales and the marketed
surplus?
Conclusions
• Rising income dampens commercialization if household
is self-sufficient
• Devaluation (or depreciation) increases
commercialization by shifting incentives toward export
crops
• Distance to market and high marketing costs reduce
commercialization by reducing fertilizer use & yields and
lowering farm gate price of commercial crops
• More…
Conclusions
• If household is net buyer of food, high marketing costs
reduce commercialization by diverting land from
commercial crops to food crops
• Higher productivity in staple food crop contributes to
commercialization by making land available
• Higher maize prices have no effect on self-sufficient
households, which explains inelasticity of supply of
staple foods
Resources
• Singh, I., L. Squire, & J. Strauss. 1986. Agricultural household
models: Extension, application and policy. Baltimore: Johns
Hopkins University Press.
• de Janvry, A., M. Fafchamps ,and E. Sadoulet. 1991.
Peasant household behavior with missing markets: some
paradoxes explained. Economic Journal 101 (409): 1400-17,
November.
• Taylor, J. 2003. Agricultural Household Models: Genesis,
Evolution, and Extensions . Review of Economics of the
Household, Vol. 1, No. 1.
http://www.reap.ucdavis.edu/research/Agricultural.pdf
• Renkow, M. Agricultural household models. Lecture notes.
http://www.agecon.ncsu.edu/faculty/renkow/ECG_540/05_household.pdf