Conservation Auction in Sumberjaya watershed:Lessons learned for Payment for Sedimentation Reduction Schemes in Indonesia

Beria Leimona (World Agroforestry Centre)LBeria@cgiar.org

Innovations for equity in smallholder PES: bridging research and practiceFriday 21st March 2014, 9.30am - 8pmCaledonian Hall, Royal Botanic Gardens, Edinburgh

A PES procurement contract auction increases

efficiency of PES contract

allocation.

Design and administer procurement auction for farmers in developing countries’ contexts

1 Auction and Contract Design

Research Site: Sumberjaya District

DamWay Besai River

55,000 ha sub-district coinciding with Way Besai upper watershed

About 40% protection forest and 10% national park

2003: 82,453 people

Density: 150 people/km 2

Coffee cultivations: monoculture and multistrata

Community as ‘land managers’

Agroforestry system (shade coffee & fruit trees) could maintain watershed function

Current public investment scheme: land rehabilitation and ‘community development’ program

Reward transfer for reducing sedimentation from upstream of the watershed

Sumberjaya: from conflict to payment for watershed services

Land conflict between local community and government: most of the land classified as protection forest and natural park

1990’s: ICRAF and local NGOs: Negotiation Support System

2003: RUPES Rewarding Upland Poor for ES

Three schemes in RES: Conditional land tenure through a modified government’s

programme on Community Forestry (temporary 5 years with the extension up to 25 years)

River Care: group contracts for sedimentation reduction.

The downstream hydropower company as the buyer with mini-hydropower as the reward.

Financial payment for individual contracts at farmers’ plot for reducing sedimentation

Developing an individual-contract scheme for reducing sedimentation using auction

Soil conservation activities

Sediment pits: 300 per hectare, standard dimensions size: 100x150x40 cm evenly distributed

Ridging: 50 percent of plot Vegetation strips: surrounding pits and

ridging Maintaining all the land conservation

structure above for a year.

Payment schedule

50 percent at inception; 50 percent at one year contingent on performance

Duration and monitoring

One year with monitoring every three months; termination if 50% contracted activities not completed by midterm monitoring date

Cancellation or non-compliance results in: ineligibility for second payment installation friction and conflict among community members indication of corruption

Force majeur provision for contract terms in the event of natural disasters

Conservation Contract

Baseline characteristics

Characteristics Natural-field auction

Number of participants 82

Distance from plot to road (mean) 25.96

Asset (mean – Rp 10 000) 7830

Education (yrs) 5.76

Current soil conservation Winners; losers

79% ; 66%

Slope (participants on slope >25%) 46%

History with external organizations(supported by at least 2 organizations)

92%

Supply curve resulting from reverse auction

Natural-field auction

2 Natural Field Auction and PES experiment

• Total participants from 2 villages: 82 farmers bidding on 70 hectares

• Participants received contracts for soil conservation: 34 farmers on 25 hectares

• Average price of contract: USD 171.70 per hectare yearly – labor requirements of contract based on

wages approximately USD 300 – Past investment for soil conservation

activities from survey USD 225

Average village compliance

Contract compliance

N % %

activities

completed

Mean bid

(Rp 10 000)

Mid-point

Non-compliant 2 6 26.13 100.00

Compliant 32 94 99.81 116.41

Final

Non-compliant 15 44.12 81.92 118.33

Compliant 19 55.88 132.68 113.16 Problem with the auction design? Good design but not understood? Optimal but realized cost and values makes it logical for people to be

aggressive early then defect

Variable Frequency

Non-contracted (N=48)

Frequency

Contracted (N=34)

Total

Understanding of the auction rules Not understand at all 3 (0.06) 0 (0.00) 3

Moderately understand 16 (0.33) 8 (0.24) 24Quite understand 11 (0.23) 12 (0.35) 23

Understand 3 (0.06) 3 (0.09) 6Understand very well 15 (0.31) 11 (0.32) 26

Complexity of the auction rulesVery difficult 2 (0.04) 1 (0.03) 3

Quite difficult 17 (0.35) 14 (0.41) 31Quite easy 7 (0.15) 10 (0.29) 17

Easy 18 (0.38) 7 (0.21) 25Very easy 4 (0.08) 2 (0.06) 6

Fairness of the auction implementation

Not fair 7 (0.15) 3 (0.09) 10Fair 41 (0.85) 31 (0.91) 72

Farmers’ understanding of auction design

Variable Frequency

Non-contracted (N=48)

Frequency

Contracted (N=34)

Total

Awareness of competition among participants

Not aware 10 (0.21) 9 (0.26) 19Aware 38 (0.79) 25 (0.74) 64

Contract value receivedToo low 19 (0.40) 5 (0.15) 24

Not too low 17 (0.35) 17 (0.50) 34Moderate 12 (0.25) 12 (0.35) 24

High - - -Too high - - -

Willingness to change the offer Yes 12 (0.25) 12 (0.35) 24No 36 (0.75) 22 (0.65) 58

Farmers’ understanding of auction design

Perspective of non-contracted and contracted farmers on social impacts

Variable Frequency

Non-contracted (N=48)

Frequency

Contracted (N=34)

Fisher’s exact test

P-value

Impact on relationships between winners and losers

0.143* 

Very bad 0 (0.00) 0 (0.00)  Bad 5 (0.10) 6 (0.18)  

Quite good 17 (0.35) 9 (0.26)  Good 21 (0.44) 19 (0.56  

Very good 5 (0.10) 0 (0.00)  Impact on general interpersonal relationships among the community

0.175 

Very bad 0 (0.00) 0 (0.00)  Bad 3 (0.06) 2 (0.06)  

Quite good 18 (0.38) 13 (0.38)  Good 21 (0.44) 19 (0.56)  

Very good 6 (0.13) 0 (0.00)  Impact on information exchange between farmers

0.055** 

Very bad 1 (0.02) 0 (0.00)  Bad 7 (0.15) 0 (0.00)  

Quite good 19 (0.40) 17 (0.50)  Good 13 (0.27) 14 (0.41)  

Very good 8 (0.17) 3 (0.09)  

Variable Frequency

Non-contracted (N=48)

Frequency

Contracted (N=34)

Fisher’s exact test

P-value

Awareness of soil and water conservation

0.188

Very bad 0 (0.00) 0 (0.00)Bad 2 (0.04) 1 (0.03)

Quite good 30 (0.63) 16 (0.47)Good 7 (0.15) 12 (0.35)

Very good 9 (0.19) 5 (0.15)Willingness to implement soil and water conservation

0.340 (0.509)

No 2 (0.04) 0 (0.00)Yes 46 (0.96) 34 (1.00)

Perspective on environmental impacts from non-contracted and contracted farmers

Note: results from 2-sided Fisher’s exact test are in parenthesis. The others are calculated from 1-sided Fisher’s exact test. For the frequency column, proportion is in parenthesis

Adopter and non-adopter

N % Mean bid (Rp 10,000)

No adoption 17 35.42 336.87

Adoption 31 64.58 388.10

• Compliance/adoption on bids is not significant

• Changes in bidding (average adjustment) not

significantly predict compliance or adoption

Can it work in a rural context of development countries?

• The auction for the PES programme in Indonesia was designed for fairness reasons. – uniform pricing – However, uniform pricing is relatively less cost-

effective compared to the discriminative price rule.

• The auction was a multiple round consisting of eight rounds with the last binding round. – Farmers learned from the rounds of the auction.

However, the announced last round may introduce forms of strategic behaviour.

– By announcing the last round, the benefits from farmers’ learning on the previous round and the advantages of a one-shot auction for the last round were combined.

Discussion• The rate of accomplishment at the final

monitoring was moderate. – lack of leadership and coordination among farmer

group members, – difficulty in finding grass seedlings to accomplish

the contract, and – coincidence with coffee harvesting time.

• In this specific case, private contract tends to be more successful compared to collective contract when leadership is lacking or “champion” among the community members does not exist. – Institutional aspects and contract flexibility might

influence the accomplishment of conservation efforts.

– Analysis showed that there were no significant differences in level of understanding, complexity, and competitiveness and conservation awareness between compliant and non‑compliant farmers.

What are factors induced a high accomplishment rate?

• A limitation of this study is that all units of the pilot site were treated as homogeneous, with respect to their contribution to erosion and downstream sedimentation.

• For a larger scale allocation auction, modifications such as using supply curve information resulting from this procurement auction would be more appropriate. – a reasonable platform for designing a scaled up

payment scheme, including differential rates and eligibility rules necessary for targeting participants.

Discussion

The design of an experimental auction should fit the purpose of overall objectives of a conservation program.

Design and administer a “fair” auction for farmers with low formal education, prone to social conflicts, and influenced by power structures within their community.

Adjustment needed for group/collective contract where decision-makers are more than a single farmer. Might not capture each individual’s preference

or true value Biased toward dominant individuals influence

contract performance

How to make it work?

Thank You

• Further information:Beria LeimonaLBeria@cgiar.org

• This research is supported by the Economy and Environmental Program for Southeast Asia (EEPSEA) and the World Agroforestry Centre (ICRAF-SEA)