A pan-tropical assessment of REDD+ socioeconomic impacts on smallholders

What impact did REDD+ interventions have on household income and inequality?

Claudio de Sassi, Amy E. Duchelle, Amy Ickowitz, Auria DP Andina, Demetrius Kweka, Erin Sills, Pamela Jagger, Sven Wunder and William Sunderlin

[email protected] ATBC - Montpellier, France – 23 June 2016

mailto:[email protected]

THINKING beyond the canopy

Concerns about REDD+ impacts on income and equity

Concerns based on theory and previous experience in conservation and development (Boerner 2011,West 2006, Ferraro 2001)

Many risks for local communities involved in global processes (Agrawal 2011, Phelps 2010)

Calls for REDD+ to appropriately compensate the full range of economic, social and political costs (Ghazoul 2010, McDermott 2013)

Empirical evidence basing on detailed income and wealth data is scarce


16 sub-national REDD+ initiatives 4000 households (randomly sampled) 125 villages (63 intervention, pre-matched to suitable control) Baseline survey before initiative implementation on the ground Impact survey ~ 3 years after initiative start Quasi-experimental counterfactual design BACI

Data: Global Comparative Study on REDD+

Comparison (Control)

REDD+ site(Intervention)

Before After

IMPACTIntervention

After

ControlAfter

InterventionBefore

ControlBefore

2010 / 2011 2013 / 2014


Data: Global Comparative Study on REDD+

Household income data includes detailed values for• Agricultural production• Forest products• Wage labor, business income, remittances, etc.• Cash and subsistence

Statistically matched to ensure balance on:• Household socioeconomic status indicators• Household demographics• Household forest uses• Household involvement in other conservation/development

programmes• Village characteristics and location


Research question 1

How does income change over time, and how does REDD+ implementation affect this change?

Household income: before-after***

***

*

*** ***

• Increase: Brazil, Indonesia, Tanzania • Stagnant: Vietnam• Decrease: Peru, Cameroon

The influence of REDD+ is modest compared to underlying income trajectories

Household income: before-after/ control-intervention

**

Before - ControlBefore - InterventionAfter - ControlAfter - Intervention


**Differences between REDD+ and control households remain constant in time, with exception of Cameroon

**


**Differences between REDD+ and control households remain constant in time, with exception of Cameroon

**

ATT


Research question 2

Do results change depending on the scale of analysis?

REDD+ impact: distribution of ATT across sites

Greater variation between sites in Brazil than other countries, comparison focused on Brazil and Indonesia (same sample size)

At the country level, only Cameroon shows an overall negative impact.

However, at the initiative level, we find that

• 5 of 16 sites show an increase in income over time, not related to the initiative

• 2 of 16 sites show a decrease in income over time, not related to the initiative

• 8 of 16 (50%) sites show no change in income• 3 of 16 sites show a negative impact of REDD+ on

household income

• However, distributional effects within communities could be masked

• No evidence using household-level data to examine impacts of forest interventions on inequality?

Results: country vs initiative level


Research question 3

Does REDD+ affect existing inequality within communities?

• Analyses of opportunity costs show widely different implications between “rich” and “poor”:• Different activities• Different per hectare income • Different shares of income affected by REDD+

Ickowitz et al, in review

• For REDD+ to fulfill safeguards and be equitable, attention on balance between benefits and burdens is needed

Income inequality in REDD+

Income inequality: Gini coefficient

Using the Gini-coefficient as a measure for income inequality we observe:

• 2 of 16 sites show an increase in inequality over time, not related to the initiative

• 13 of 16 sites show no change in inequality• 1 of 16 sites show a negative impact of REDD+ on

inequality

Before After

High income HH

Low income HH

“NO HARM” REDD+: does not affect underlying trendsH

ouse

hold

inco

me

Before After

High income HH

Low income HH

PROGRESSIVE REDD+: reduces existing inequalitiesH

ouse

hold

inco

me

Before After

High income HH

Low income HH

REGRESSIVE REDD+: accentuates existing inequalitiesH

ouse

hold

inco

me

Before After

CONTROL INTERVENTION

Before After

Income trajectory by wealth group: Brazil

Wealth quintileslowhigh

Before After


Before After

Income trajectory by wealth group: Brazil

Wealth quintileslowhigh

Difference BACI

Difference BACI

Income trajectory by wealth group: Cameroon (1 site)

lowhigh

Wealth quintiles

Before After


Before After

Increase in income inequality as the poorest people (quintile) is more negatively affected than the rich

Income trajectory by wealth group: Cameroon (1 site)

lowhigh

Wealth quintiles

Before After


Before After

*

Difference BACI

Difference BACI

Increase in income inequality as the poorest people (quintile) is more negatively affected than the rich


Conclusions REDD+ impact on household income is small in size

compared to background dynamics

No evidence so far of REDD+ increasing income as social co-benefit

Some evidence of REDD+ negatively affecting income and/or increasing inequality at some sites should serve as an important warning

Conclusions on pooled data overlook important site-level differences and belie dynamics related to pre-existing inequality and wealth distribution

Financial support for GCS-REDD+:Norwegian Agency for Development Cooperation, Australian Agency for International Development,

European Commission, UK Department for International Development,

German International Climate Initiative,CGIAR Forests, Trees and Agroforestry (FTA) Programme.

www.cifor.org/gcs

http://www.cifor.org/gcs

A pan-tropical assessment of REDD+ socioeconomic impacts on smallholders

Environment

Transcript of A pan-tropical assessment of REDD+ socioeconomic impacts on smallholders