Designing Payments for Environmental Services in the Context of Weak Property Rights and Commercial...

Designing Payments for Environmental Services in the Context of Weak Property

Rights and Commercial Interests

Stefanie Engel and Charles Palmer

Center for Development Research (ZEF), University of Bonn

ZEF Bonn

Funding: Robert Bosch Foundation

Indonesian Setting

• 10 % of world tropical forests

• Tremendous biodiversity

• 40% of forests lost in 1950-2000

• Rate accelerating (2 mio ha/year)

• Lowland tropical forests predicted to disappear in Sumatra and Kalimantan by 2010

• Commercial exploitation as a main cause

Indonesian Setting(cont.)

• Post-Suharto Decentralization Reforms

• Acknowledgement of communities‘ forest ‘rights’

• Exercise rights by negotiating logging agreements

• ‘Selling off forests for a satellite dish’

• Payments for Environmental Services (PES) could provide alternative income source

Issues in designing PES

1. Effectiveness

• How much to pay?

• Conventional answer: Opportunity costs = expected payoffs from logging deals

• Additionality, Enforcement

Community payoffs from negotiated agreements in Indonesia

Survey results from East Kalimantan, Indonesia

62 community-firm logging agreements

Actual fees paid (US$ per m3):

Min=0.28, Max=11.81 =3.54

Actual fee paid to community plus value of social developments

Freq

uenc

y

10

8

6

4

2

0

Std. Dev = 24131.25

Mean = 32788.8

N = 62.00

Issues in designing PES

1. Effectiveness

• How much to pay?

• Conventional answer: Opportunity costs = expected payoffs from logging deals

• Additionality, Enforcement

Issues in designing PES (cont.)

2. Efficiency

• How to maximize environmental services with given budget?

• Conventional answer: Target communities with lowest expected payoffs from logging agreements

Complexities in our setting

• Opportunity costs are result of community interactions with commercial interests (logging companies)

• Strong variation in observed logging fees received

• Weak property rights -> Can communities really protect forest?

=> Conventional answers are misleading!

Game-theoretic model

– Simplified version of Engel, López & Palmer (forthcoming) and Engel & López (2004)

– Integrating conflict and bargaining theory

– Assumptions

• 2 actors: Community, Firm

• Only firm has access to capital

• Weak community rights over forest

• Perfect information

• Risk neutrality

A simplified game-theoretic model- Component 1: Negotiation

Alternating-Offers Bargaining GameNet LoggingProfits (v-c)

Community payoffs (C) = CC R,~

max

= dC + (v – c – dC – dF) (Nash Bargaining Sol.)

= Bargaining power of C relative to F (0 1) (depends on discount rates and other factors)di = Inside option of actor i (i=F, C) Ri = Outside option of actor i

C~

Here: Inside and outside options depend on who can claim de facto rights over forest if negotiations do not succeed

A simplified game-theoretic model- Component 2: Conflict over de facto rights

Fwithdraw

Attemptlogging

C

withdraw

Establishblockade

withdraw

Maintainblockade

F

F……….

C

withdraw

withdraw

t=0 t=2t=1……….

Attemptlogging

Attemptlogging

WAR OF ATTRITION (based on Burton, ERE 2003)

Costs: s (per period)

Benefit in case of winning: Value of standing forest b (per period)

Discount rate: rC

Costs: c (per period)

Benefit in case of winning: Profit v -

Discount rate: rF

Cmin

brCs

Firm wins potential conflict->Logging takes placeCommunity receives

Unprofitable logging

Community wins potential conflictLogging negotiations succeedCommunity receives

Community wins potential conflictLogging negotiations fail-> Forest conservation

I

III

II

v

Cv min

CF RRcv

Cc min

s

Cmin

CC R,~

max

FC

CC

CF)r1ln(

)r1ln(C

Fd

rC

bcv

r

b~,

r

bR,r1

bs

)r1(s

r

c C

F

Logging profits

C‘s valuationof standingforest

Implications for PES Design

• PES introduces new value of standing forest to community:

b = b0 + P

C‘s valuationof standing forest per period without PES(fuelwood, non-timber forest products,hunting,local ecological services)

PES amount per period

br C s

F i r m w i n s p o t e n t i a l c o n f l i c t- > L o g g i n g

U n p r o f i t a b l e l o g g i n g

C o m m u n i t y w i n s p o t e n t i a l c o n f l i c tL o g g i n g n e g o t i a t i o n s s u c c e e d- > L o g g i n g

C o m m u n i t y w i n s p o t e n t i a l c o n f l i c tL o g g i n g n e g o t i a t i o n s f a i l- > F o r e s t c o n s e r v a t i o n

I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

PES may not be enforceable

Logging profits


=b0 + P

=> PES ineffective

br C s





I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

PES raises C‘s ability of enforcement, improving logging deal

Logging profits


=b0 + P

=> PES ineffective

br C s





I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

PES raises C‘s outside option, leading to better logging deal

Logging profits


=b0 + P

=> PES ineffective

br C s





I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

Effective PES

Logging profits


=b0 + P

Conditions for effective PES

1. Community needs to be able to win potential conflict with the firm (otherwise firm can log despite PES agreement)

2. PES needs to induce negotiation failure

F)r1ln(

)r1ln(

0

C

FCmin0 r1

Pbs

)r1(s

r

cwhere,)Pb(v

C

F

0FC bRcvrP

Maximum that Fwould ever offer(F‘s surplus)

C‘s ownvaluation of forest

C‘s discount rate

br C s





I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

Case 3: No additionality

Logging profits


=b0 + P

br C s





I

I I I

I I

v

Cv min

CF RRcv

Cc min

s

Efficiency of PES

Logging profits


=b0 + P

Conditions for effective PES

1. Community needs to be able to win potential conflict with the firm (otherwise firm can log despite PES agreement)

2. Needs to induce negotiation failure

0FC bRcvrP

F‘s surplus(maximum itwould ever offer)

C‘s ownvaluation of forest

C‘s discount rate

Poverty-EnvironmentTradeoff!

F)r1ln(

)r1ln(

0

C

FCmin0 r1

Pbs

)r1(s

r

cwhere,)Pb(v

C

F

br C s

I m p r o v e d l o g g i n g d e a l s


L o g g i n g n e g o t i a t i o n s s u c c e e d- > L o g g i n g

L o g g i n g n e g o t i a t i o n s f a i l- > F o r e s t c o n s e r v a t i o n

I a

I I I

I I

v

Cv min

CF RRcv

Cc min

s

I b

A d d i t i o n a l c o n s e r v a t i o n

What if property rights are secured?

Logging profits


=b0 + P

Empirical application

• Use fieldwork data to estimate boundary conditions:

a. Probability of winning conflict

b. Probability of negotiation failure

• Collect data on potential PES communities

• Compute predicted probabilities a. and b.

• Choose communities with one of these probabilities above 0.5 and the other below, but close to 0.5

Variable (Parameter proxied)

Coeff Std. Error

Constant -3.2680 2.6455 Average proportion of household incomes derived from sale of forest products, AVINCFP (b)

0.1587 0.2272

Forest quality: area logged before by commercial operation (Yes=1, No=0), LOGGEDB4 (v)

2.1276 1.1978*

Actual area logged (ha), ACTHA (v) -0.009559 0.004437** Proportion of households containing at least one government employee, PGOVJOB (s)

-0.7598 0.5081

Proportion of households that participate in community organizations, HPARTORG (s)

0.7434 0.3270**

Proportion of households containing members of dominant ethnic grouping, PHHDOMGP (s)

0.3034 0.2534

Distance to nearest market (km), DISTMARK (s) -0.2566 0.1570 Proportion of hhs holding savings before agreement, HBANKACC (rC)

0.1267 0.5515**

No. of observations 62 % of outcomes predicted correctly 71%

Empirical estimation: Probability of winning conflict

Probability of receiving payoff above minimum negotiated level (15.000 IDR)

Empirical application

• Use fieldwork data to estimate boundary conditions:

a. Probability of winning conflict

b. Probability of negotiation failure

• Collect data on potential PES communities

• Compute predicted probabilities a. and b.

• Choose communities with one of these probabilities above 0.5 and the other below, but close to 0.5

Empirical application (cont.)

• Computation of minimum payments difficult

• Alternative:

– Use previous analysis to assure additionality (starting point in area I or II)

– Use auction/contract design to let communities self-reveal opportunity costs (size of payment required)

Conclusions

• Conventional wisdom:

– Estimate expected logging payoffs based on observed variation

– Pay at least these expected payoffs

– Focus on communities with low expected payoffs

• Incorrect here

– Communities with lowest expected payoffs cannot enforce forest conservation

Failure to consider weak property rights leads to lack of enforceability

– Introduction of PES raises expected logging payoffs

Failure to consider endogeneity of payoffs leads to better logging deals rather than conservation

Conclusions (cont.)

• PES design as complex task

• Need to compensate for best possible offer by company (unobserved) minus C‘s own valuation of standing forest

• Poverty vs. Conservation Tradeoff

• Applicability to other situations with weak property rights and commercial interests

Additional slides

Outline

1. Introduction

2. Game-theoretic model of community-firm interactions

3. Implications for PES design

4. Empirical application

5. Conclusions

Conflict outcomes

Cmin

F)r1ln(

)r1ln(C

Fr1

bs

)r1(s

r

c C

F

brcs

•F wins potential conflict•Logging•C‘s payoff =

Unprofitable logging

•C wins potential conflict•C gets present value of standing forest

(b/rC)•Negotiations possible

vCv min

cc min

s

cmin

Logging profits


So, what determines community payoffs?

Cmin

F

CC

Cmin

Cmin

C

,...)b(vifdr

bv1

r

b

,...)b(vif

Combining conflict and bargaining results...

Depends on:

• Value of standing forest (b), bargaining power, profitability, etc.

• Ability to self-enforce property rights (win conflict) Depends on value of standing forest (b), other parameters

Moreover: Negotiations fail only if v – c < RC + RF

( Size of cake < sum of outside options

Determinants of Community Payoffs - Hypotheses

Determining factors Effect in bargaining

Effect on property rights

Total

Logging profit (v) + - ?

Fixed logging costs (c) - + ?

Community‘s valuation of standing forest (b)

+ + +

Community blockade costs (s) Kein Effekt - -

Firm profits in next-best activity (dF) - Kein Effekt -

Community discount rate (rC) - - -

Firm discount rate (rF) + + +

Other factors increasing community bargaining power (z)

+ Kein Effekt +

Empirical Analysis – Proxies from Indonesian Survey

FACTOR PROXIES

Logging profits and costs (v, c)

Size of concession area; Dummy, whether area was logged before

Community‘s valuation of standing forest (b)

Average % of household income from forest products

Community blockade costs (s)

% of households (HH) with at least one government employee; Distance to market; %

of HH with members in community organizations; % of HH belonging to ethnic

majority;

Firm profits in next-best activity (dF)

NA

Community discount rate (rC) % of HH with savings (bank account) before negotiations

Firm discount rate (rF) NA

Other factors increasing community bargaining power

% of HH with previous experience working for a logging company

Variable (Parameter proxied)

Actual logging fees received Actual logging fees + social developments

Coeff Std.Error

t-stat. P[|T|>t] Coeff Std. error

t-stat. P[|T|>t]

Average income from forest products

263.9940 153.60

1.719 0.0915* 270.6891 152.13

1.779 0.0809*

% households with logging experiences

157.7643

112.26

1.405 0.1657 154.2934

111.68

1.382 0.1729

Actual area logged -2.860419

0.6420

-4.456 0.0000**

-2.933886

0.6430

-4.563

0.0000**

Forest quality (logged before)

7943.076

6022.8

1.319 0.1929 7827.253

5936.3

1.319 0.1930

% households with government job

-507.9924

288.51

-1.761 0.0841* -439.3044

305.97

-1.436

0.1569

% households that participate in village

156.5899

78.664

1.991 0.0517* 143.2843

77.722

1.844 0.0708*

% households from dominant ethnic gp.

139.7951

76.568

1.826

0.0735* 151.2842

76.619

1.975 0.0535*

Distance to market -176.0127

51.056

-3.447 0.0011**

-171.2411

49.821

-3.437

0.0012**

% households with bank accounts/savings

358.9818

174.55

2.057 0.0447**

356.4649

172.97

2.061 0.0442**

0.44348 0.44588Adjusted R-squared

Empirical Analysis – Econometric Results

Empirical Analysis - Results

• Confirms hypotheses derived from theoretical model

• Empirically significant factors:– Importance of forest for household incomes (+)

– Size of concession area (-)

– Household income (Bank account) (+)

– Determinants of collective action (Blockade costs):

• Opportunity costs (government jobs, distance from mkt.) (-)

• Social capital (Participation in organisations) (+)

• Social homogeneity (% in ethnic majority) (+)

Community payoffs depend crucially on (potential) ability of the community to self-enforce its property rights over the forest (win a conflict).

Implications for PES Design (cont.)

Introduction of PES raises community‘s valuation of standing forest

Raises community‘s ability to win potential conflict and its inside/outside option in bargaining

Raises community‘s expected payoff from negotiations because firm would offer more

If this endogeneity is ignored, community may simply negotiate better logging deal

Designing Payments for Environmental Services in the Context of Weak Property Rights and Commercial...

Documents

Transcript of Designing Payments for Environmental Services in the Context of Weak Property Rights and Commercial...