Research and Development Roundtables (RRD in French) on the theme: « Fixing Agriculture Value Chains»
1. Quality certification and staple crop
value chains in Sub-Saharan Africa
Tanguy Bernard
Rencontres Recherche et Développement
IFPRI – Dakar, 30 Mai 2018
2. Fact #1: Production of staple crops is high on African policy agenda
- 2008 food crisis has revealed vulnerability of
African countries to movements in international
prices
- Historical specialization on export crops
(coffee, cocoa, groundnuts, cotton, etc.)
- Increased dependence on imports over past
decades
- Since 2008: trade and production policies to
support local production of staple crops
- Staple crops must compete with imports.
- Quality increasingly demanded by processors
and urban consumers
SSA net exports for selected commodities (FAOStat)
3. Fact #2: African agriculture is small-scale
- Structural transformation has not yet led to
consolidation of farm-land.
- Issues with land markets (cf. titling programs)
- Incomplete rural-to-urban migration
- Increased rural population pressure
4. Fact #3: Limited agricultural technology adoption by small farmers
• Multiple constraints
• Supply side issues: lack of local access to
technology
• Demand side issues: lack of knowledge, time
inconsistency
• Contextual issues: lack of financial services:
credit, insurance, payments
• But low expected return likely determinant of low
adoption
Source: Sheahan and Barrett, 2017
5. Fact #4: Transactions are small, aggregation is necessary
02468101214161820
Percent
0 <=1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 >10
Monthly transactions per household (quintals)
- Due to small farmsize
- Because small farmers sell in small quantities at once:
- Exploit seasonal price variations
- Favor home consumption as protection from
market variations
- Stored harvest more difficult to spend (discipline)
- Some heterogeneity
- Cash vs staple crops
- Husbands vs wives (responsible for spendings of
different frequaence and magnitude)
- Necessary aggregation at local level
Wheat transactions in Ethiopia, 2016
6. Typical organization of a staple crop value chain in Africa:
A market for lemons
Aggregators
?
coop trader
miller
= =
(few contracts)
7. To what extent can better quality
recognition help local producers
compete with imports?
8. Quality certification on smallholder farmers three main
commercialization outlets
• Contracts :
• Cooperatives/producer groups
• Local rural markets
10. Fact #5: small producers sell locally
• High fixed transport costs reduce possibility for spatial
arbitrage
• Local markets are thin, yielding to important price
fluctuations.
• Access to price information has little effect on
producers’ prices
• E.g. Aker and Fafchamps in Niger
• Traders may collude (mixed evidence)
• E.g. High pass-through: Casabury
• E.g. Low pass-through: De Falcqo, Fafchamps and
Hill
Daily mean price on 3 onion markets in Podor department, Senegal
11. Fact #6: Quality is weakly recognized at transaction-level
• Grades and standards are largely inexistent
at small-scale level.
• Quality appreciation (and often weight)
is part of the price negociation.
• Uncertain returns to quality may
disincentivize producers to enhance the
quality of the products.
65707580859095
400 500 600 700 800 900 1000 1100 1200 1300 1400
bandwidth = .8
12. • Basic daily ingredient in every Senegalese
kitchen
• Mostly imported from Holland
• Low quality makes domestic onions not
competitive with imports
• A huge policy issue as delinks domestic
producers from domestic consumers
• Policy response:
• Since 2000, 7 months import ban to encourage
domestic production
Onions in Senegal
13. Onion production in Podor
• Important production zone: Podor department
(study area) 3,500 ha
• Strong regional development agency (SAED)
• Large access to inputs through dense network of retail
shops
• Distant to markets, but good infrastructure
• Size and quality of onions depend on fertilizers
used
• Use urea to produce larger onions to fill bags, but high
water intensity and high perishability (high post-harvest
losses)
• Use 10-10-20 N-P-K to increase weight and quality
14. • Farmers bring product to consignment agents
(coaxers) on local assembly markets
• For fixed fee, coaxers negotiate and sell to
long-distance traders (banabanas)
• Onions sold on volume in presumed 40 kg
bags
• No scales or quality measures as “banabanas
would no longer come”
Onion commercialization in Podor
15. Onion value chain in Senegal
The organization of the value chain is such that producers are incentivez to produce LOWER QUALITY onions
?
Itinerant
traders
(banabana) on
rural assembly
markets
Final consumers
=
16. • Policy context
• 2013: Local market authorities agree to the introduction of scales and quality labeling (three
grades)
• 2014: We work on implementation of weighing and labeling with local University Gaston Berger
• Study design
1. Universal training for onion producers on quality enhancing technologies and practices (SAED)
2. Information campaign on scales/labels in random half of 34 villages delivering onions to assembly
markets Assess effect on production and marketing behavior
3. Use delays in authorization for effective operation of scales as a time discontinuity for Diff-in-Diffs
Assess effect on prices for farmers in treated vs. control villages before and after
Experimental design
17. Time line of experiment
Enable quality response in all villages (training)
Induce production responses for quality in treated villages (information campaign)
Induce marketing response (sorting for quality): can be done by all at market
Observe price effect by double difference for treated villages when labeling effective
18. Information campaign and knowledge about scales
Ultimately, all farmers knew
about introduction of scales.
But farmers in T villages
learned about it early on
(January) through the
information campaign vs. at
delivery for C farmers
Farmers in T villages had
time to adjust their
production and marketing
practices
19. Results: Impact of market reform on production and
marketing behavior
Information about market reforms induced a change
in production behavior:
• 9%pts decline in incidence of use of pro-volume urea from a
base of 95%
• 27%pts increase in incidence of use of pro-weight/quality 10-
10-20 from a base of 28% (doubles)
• Increase in application of 10-10-20 by 116 kg/ha from a base
of 43 kg/ha (nearly triples)
All farmers increased sorting, especially for
transactions occurring after the introduction of scales
and labels.
• Does not require information about the reforms (T) as can be
done upon arriving at the market
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
% use Urea % use 10-10-20
Control villages Treatment villages
20. Results: Impact of reform on quality and price
Impact on quality
• 16%pts increase in likelihood of onions being of
good quality from a base of 8%
Impact on price
• Diff-in-Diffs before-after scales effective, for
treatment (can adjust quality as informed) vs.
control
• Test of parallel trends before introduction of
labels satisfied
• Diff-in-Diffs 6 to 9% increase in price received
after introduction of labels if informed
108
110
112
114
116
118
120
122
0 1 2 3 4 5 6 7 8 9
Control villages Treatment villages
Introduction of
labelsPrice per kg
Days
21. Results: Testing for quality as the channel to price increase
Observe partial correlation between quality
and price
• Higher price received correlates with higher
quality
• Price effect mainly associated with introduction
of labels
0
20
40
60
80
100
120
140
160
Low quality High quality High quality and labels
Price per kg
22. Increase in revenue
No change in number of bags harvested per ha
Increase in weight of bags: quantity effect
Increase in quality of onions sold: quality effect
Increase in price received if informed: +6 to 9%
Increase in cost
Fertilizer, sorting
Increase in net income per hectare: +11%
Reverse “market for lemons” effect
High quality sold through certification system
Low quality sold directly on volume through coaxers
Cost and benefit
23. • Direct effect:
• Certification led to higher
prices ofr better quality
onions
To sum up
• Indirect (most important) effect:
• Existence of certification scheme has encouraged
farmers to chnage their production practices towards
higher quality onions
24. Epilogue: despite the benefits, use of scales and labeling was
discontinued at end of experiment
Some evidence that traders
lobbyed against the system
- Likely do not perceive
increase in quality of onions
Market reform requires
collective action or
government intervention
25. • Market reform led to increased adoption of quality-enhancing technology in
production and of sorting by quality in marketing
• Role of market information on quantity and quality :
• Increase in price premium received by farmers on assembly markets: 6-9%
• Some increase in weight
• Overall net benefit to farmers: 11% increase in net income per hectare
• Hence, some price pass-through to farmers achieved
• Importance of farmers’ behavioral response in overall benefit from innovation.
Behavioral response can be a large share of total gain from innovation: need
be understood, facilitated, and amplified
Summary of results
27. • Contract enforcement issues when:
• Opportunities for side-selling are
important (active local market for the
crop)
• Size of transactions are small which
makes monitoring or legal recourse
too expensive
% farmers under contract-farming arrangements
Source: from Minot and Sawyer, 2016
Fact #6: Contract farming is rare and limited to cash crops
28. Related study : Independent quality certification in contract
farming arrangements
(Dairy in Vietnam (Saenger, Torero & Qaim; 2014))
- Milk is bulked at collection point
- Samples are taken from each producer
- Company analyses samples and pays farmers
according to the quality they delivered.
Issue: farmers cannot observe quality, limited trust in
company’s quality assessment
Producers are given vouchers to test for their milk
quality in an independent laboratory, if they want
to contest the company’s quality assessment
results.
Collection point
Dairy
Control group
Collection point
Dairy
Treatment group
29. Results
• Main effects
• No evidence of company’s misreporting of milk quality
• Changes in production behavior amongst producers in Treatment: increase input use, increase
quality, increase output.
• Positive net income effect for producers
• Conclusion
• the buying company did not behave opportunistically, but the supply chain architecture did not
allow the buyer to signal its fair type to farmers.
• Third-party enforcement of contracts in an environment of weak institutions can move the supply
chain to a first-best scenario in which both smallholders and buying companies benefit from
increased farm productivity.
31. Fact #5: Collective commercialization functions poorly
• Before structural adjustment plans, many
value chains organized with government-
controled cooperatives.
• Farmers had little choice but to sell
through cooperatives
• Cooperatives and farmer groups
largely spread throughout the
countries
• SAP: cooperatives are seldom able to
organize collective commercialization.
• With some exception in a few cash-
crop value chains
% Villages
with at least
one market-
oriented
RPO
Within these
villages, % hh
that are
members
% RPOs
effectively
active in
output
commercializat
ion
Sénégal (2002) 47 75 38
Ethiopia (2006) 56 45 59
Burkina Faso (2002) 35 17 59
Ghana (2011) 31 15 37
32. • Experiment
• Trained farmers to measure unobservable
quality of their wheat (flour extraction
rate). Compare to control group.
• Result
• Vary with quality of output
• Farmers with higher output quality:
• Increase share of wheat sold
• Sell less through local cooperatives (which
do no differentiate quality)
• Receive higher price
• Increase investments in quality-
enhancing fertilizer in later season.
Related study : Quality measures and supply to cooperatives
(Wheat in Ethiopia (Abate & Bernard; 2017)
30.0035.0040.0045.00
Percentageshare
share sold low high remote less-remote
grain quality remoteness
Control Quality trg
Share of wheat sold (marketed surplus)
33. To conclude
• Strong producers’ responses to quality recognition
• All three main outlets: rural markets, contracts, cooperatives
• Through production and commercialization related behaviors
• Evidence of positive welfare effects
• A case for producer-level G&S in Sub-Saharan Africa
• G&S virtually absent at producer-level (some exception for private standards in cash-crop
sectors).
• Absent G&S, disincentive for quality improvement at farm-level, and aggregation induces
quality losses. G&S have potential to affect all downstream segments of value chain
• Can contribute to agriculture modernization and competitiveness in Sub-Saharan Africa
• Several open questions for conditions of successful implementation
Requires careful piloting and evaluation