Balancing Global Climate Action and Local Rights: Survey Experimental Evidence on Public Support for Carbon Offsetting in Liberia

The intervention in our experiment is a vignette embedded in a survey (i.e., information treatment) designed to gauge responses to different narratives about carbon offsetting projects in a developing country context. The survey includes a common narrative presented to all participants, followed by two distinct types of narratives – one highlighting the positive aspects and the other emphasizing the negative aspects of carbon offsetting projects. In order to help the understanding of the respondents and for ease of communication, the narratives were supplemented with single-page black and white cartoon visually depicting the narrative.

1. Positive Narrative: This survey component describes the beneficial impacts of carbon offsetting projects. It focuses on how such projects can introduce financial investment from abroad and promote the adoption of cleaner, more sustainable technologies in the host country.

2. Negative Narrative: This part of the survey outlines the potential drawbacks of carbon offsetting projects. It emphasizes concerns such as the potential for these projects to encroach on the land and rights of local communities and doubts about the actual effectiveness of these projects in achieving their intended environmental goals.

2023-12-01

2023-12-31

The key outcome variables in our experiment are structured into four primary outcome clusters/indices:

1. Government and Policy Perspectives Index
2. Environmental and Investment Perceptions Index
3. Social and Community Concerns Index
4. Global Climate Change Perspectives Index

Each index represents a distinct thematic area, capturing attitudes and opinions across different dimensions related to carbon offsetting, environmental policies, and socio-economic impacts.

The primary outcome clusters/indices are constructed from the responses to eight Likert-scale questions. We will evaluate these items separately as well as as indices which are standardized into z-scores using inverse covariance weighting (Anderson 2008). This method adjusts for potential correlations between items. The standardized responses are then grouped and further standardized within their respective thematic categories to form the four clusters/indices:

1. Government and Policy Perspectives Cluster/Index: Combines scores from questions about
- support for government involvement in carbon offsetting,
- trust in government’s environmental management, and
- views on policies affecting jobs in agriculture and forestry.

2. Environmental and Investment Perceptions Cluster/Index: Includes scores from questions assessing
- perceptions of the benefits of carbon offsetting projects and
- the importance of green technologies for climate change.

3. Social and Community Concerns Cluster/Index: Derived from responses to
- concerns about local communities losing homes and land rights due to carbon offsetting projects.

4. Global Climate Change Perspectives Cluster/Index: Aggregates scores from questions on
- the effectiveness of international organizations in climate change management and
- views on climate change as a global challenge requiring international cooperation.

Our experiment employs a vignette-based survey design to examine the impact of different narratives on perceptions of carbon offsetting projects in a developing country. The design incorporates a common narrative given to all participants, supplemented by either a positive or negative narrative about carbon offsetting, or both. These narratives are cross-randomized into four distinct groups, with a total of 2,000 subjects equally distributed across these groups.

1. Group 1 (Control): Receives only the common narrative, serving as the baseline for comparison. This group does not receive any additional positive or negative information about carbon offsetting projects.

2. Group 2 (Positive Narrative): Receives the common narrative along with additional information highlighting the positive aspects of carbon offsetting projects, such as economic benefits and adoption of sustainable technologies.

3. Group 3 (Negative Narrative): Receives the common narrative and a narrative focusing on the potential negative impacts, like land appropriation concerns and effectiveness doubts.

4. Group 4 (Combined Narrative): Participants in this group are exposed to both the positive and negative narratives, in addition to the common narrative. This group is further split, with half of the subjects (250 out of 500) randomly selected to receive the positive narrative first, followed by the negative narrative, and the other half receiving these in the reverse order.

This design allows us to assess the individual and combined effects of positive and negative information on participants' perceptions, and the order effect in the group receiving both narratives.

Randomization was conducted in the field using a random number generator feature within SurveyCTO, a data collection software utilized on Android tablets.

Randomized at individual level

No

2,000 individuals

2,000 individuals

500 individuals: Group 1 (Control)
500 individuals: Group 2 (Positive Narrative)
500 individuals: Group 3 (Negative Narrative)
500 individuals: Group 4 (Positive + Negative Narrative)

Based on the power calculation with a total sample size of 2,000 participants divided equally into four groups (500 in each group), a significance level of 5%, and a desired power of 80%, the minimum detectable effect (MDE) size for our experiment is approximately 0.177 standard deviations.