Presented by David J. Spielman and Patrick Ward, IFPRI at the Africa RISING–CSISA Joint Monitoring and Evaluation Meeting, Addis Ababa, Ethiopia, 11-13 November 2013
Economic policy analysis tools for sustainable intensification
1. Economic policy analysis tools for
sustainable intensification
David J. Spielman and Patrick Ward
Africa RISING–CSISA Joint Monitoring and Evaluation Meeting,
Addis Ababa, Ethiopia, 11-13 November 2013
2. CSISA’s policy research agenda
Improved policies & institutions for inclusive growth
Aims and objectives
– Address the policy environment needed to remove constraints
to the adoption of new technologies
– Enhance the productivity and income benefits of improved
agricultural sector growth
– Explore strategies and solutions to catalyse durable change at
scale
Expected outcomes
• Improved policies and incentives that…
– encourage private investment and public-private partnerships in
pro-poor technology development and delivery
– address changing labor, gender, assets, and migration dynamics
related to pro-poor technology development and delivery
3. CSISA’s interesting questions
• What are farmers willing to pay for new technologies?
– Custom-hired resource-conserving machinery services
– Abiotic stress-resistant traits in inbred vs. hybrid backgrounds
• How heterogeneous in this demand for technology?
– Does demand differ across geography, wealth, income, social status, etc.
– Does heterogeneity create market opportunities for private investment?
• How might public and private investment leverage
demand heterogeneity to increase adoption?
– What types of encouragements (subsidy vouchers, discount coupons)
encourage adoption?
– How do social network effects encourage adoption?
– What are the tradeoffs inherent in different encouragement schemes?
4. CSISA’s policy analysis toolkit
• Tools to elicit willingness to pay and risk preferences
– Choice experiments
– Experimental games
• Tools to evaluate impact of a new technology
–
–
–
Natural experiments (difference-in-difference)
Propensity score matching (PSM), Regression discontinuity designs
(RDD), Instrumental variables (IV)
Controlled experiments (RCTs)
• Tools to evaluate market opportunity
– Industry analysis, new empirical industrial organization
– Structure, conduct, performance analysis
5. Ex: Preferences for drought tolerance in rice
• What are farmers willing to pay for drought-tolerant
rice in inbred v hybrid background?
– Inbreds: low-cost “renewable” public sector technology
– Hybrids: high cost “non-renewable” private sector technology
• Participants presented with a series of “choice tasks”
meant to reflect real-world seed purchasing decisions
• Three varying alternatives (A, B, or C): comprised of different
combinations of relevant attributes
• Status quo (D): option to use seed cultivated in previous kharif
season
Source: Ward, Ortega, Spielman and Singh (2013)
7. Estimated willingness to pay
Seed must be purchased every year
Attribute
Low seeding rate (4-6 kg/acre)
Duration: 120-135 days
Duration: <120 days
Yield: 1/3/0.6/0.4
t/acre under
normal, severe,
extreme drought
Yield 1/3/0.7/0.4
Yield: 1.3/0.8/0.4
Yield: 1.3/0.9/0.4
-100
-50
0
50
100
150
200
Indian Rupees (INR)
Source: Ward, Ortega, Spielman and Singh (2013)
8. Demand for DT inbred vs DT hybrid rice
Source: Ward, Ortega, Spielman and Singh (2013)
9. Ex: Willingness to pay for LLL
• What are farmers willing to pay for custom-hired LLL
services?
– Binding auction for LLL service at for up to 3 plots of land
– If participant bid ≥ a randomly drawn price, he may receive the service
– Auction repeated in 2011 and 2012
Source: Lybbert, Magnan, Spielman, Bhargava, and Gulati (2013)
10. Heterogeneous willingness to pay across
multiple dimensions
Source: Lybbert, Magnan, Spielman, Bhargava, and Gulati (2013)
11. Change in willingness to pay, 2011-2012
Source: Lybbert, Magnan, Spielman, Bhargava, and Gulati (2013)
12. Cost-effectiveness of targeting strategies
(Subsidy cost per acre and farmer)
350
Subsidy cost per acre leveled
(Rs./acre)
District
BPL
300
Uniform
Landholdings
250
Perfect
200
Net gain for provider
Net loss for provider
First-hour
discount
150
100
100
150
200
250
300
Subsidy cost per farmer leveling (Rs./farmer)
350
Source: Lybbert, Magnan, Spielman, Bhargava, and Gulati (2013)
13. From auction to lottery: social network effects
Auction losers
Random sample from
village v
Auction
Auction winners
(self-selection)
Lottery
(random selection)
Lottery winners
Lottery losers
Outcome:
Water savings ≈ 21%
Source: Magnan, Spielman, Lybbert, and Gulati (2013)
14. New evidence, actionable information
• We can design novel targeting strategies to promote
new technologies
– A private sector strategy: Firms promote technologies to
profitable market segments only
– A public sector strategy: State promotes technologies to all
market segments
– A synthesis strategy: Firms and state collaborate to target
beyond profitable segments only
• Each targeting strategy involves tradeoffs
– Equity vs. efficiency
– Productivity vs. sustainability