Reducing marine debris pollution by changing household behavior through children education

We design and implement a randomized field experiment targeting a sample of 15 matched pairs of schools participating in the sustainable school program, led by the Ministry of Environment of the Government of Chile (MMA). The experiment involves the provision of an environmental education program with value-laden content. We target students attending all fourth grades in the treated schools. Because the program aims at assessing both a direct and a spillover effect, environmental education is coupled with activities that involve the parents.

We design and implement a randomized field experiment targeting a sample of 15 matched pairs of schools participating in the sustainable school program, led by the Ministry of Environment of the Government of Chile (MMA). The experiment involves the provision of an environmental education program with value-laden content. The program consists of the implementation of the contents embedded in the educators’ guide to marine debris, designed by the North American Marine Environment Protection Association (NAMEPA) in partnership with the National Oceanic and Atmospheric Administration (NOAA). This guide is based on NOOA’s turning the tide on trash: A learning guide on marine debris (NOAA, 2015). We target students attending all fourth grades in the treated schools, and therefore the program include the three modules devoted to grades K-5 on this guide. Each module is covered in two lessons, and the activities are customized to the Chilean context when necessary (e.g., endangered species, locations, etc.). The lessons take place in the classroom, and the contents are taught by a certified teacher and a support team that keep detailed records of the lecture, that allow us to explore mediators that could potentially affect the delivery of the message (i.e. attendance, attitudes, participation, etc.). The duration of each lesson is two times 45 minutes, which are administered biweekly, with in total 6 lessons. Because the program aims at assessing both a direct and a spillover effect, the lessons are coupled with homework and activities that involve the parents (e.g., counting the number of single-use plastic in the home, counting and sorting the number of single-use plastic that are used during a week, etc.). The value-laden content of the environmental education program consists of personal normative messages in the lecture material and the activities, which are designed to be shown to parents indirectly (e.g., school art and craft projects). This has been proved to generate a larger impact on plastic consumption and waste separation by making salient that both the cause and consequence of marine plastic pollution problem has its root in the individual behavior. This opposes to information provided alone which is shown to have a shorter effect (Xu et al., 2018; Willman, 2015; Jakovcevic et al., 2014; Convery et al., 2007).

2019-06-17

2019-12-13

Based on the KAP module, we have three main outcome of interest: knowledge, attitudes and practices. Knowledge about plastic pollution problem is measured as the percentage of correct answers from a list of 11 questions. Attitudes regarding plastic pollution problem and disposal of plastic are proxied by an index comprising 6 questions. Finally, practices include a series of actions carried out by individuals, covering different levels along the plastic impact pathway (i.e., consumption and disposal of plastic), as suggested by Alpizar et al (2019). The outcomes of interest in terms of consumption are: (i) An index of plastic composition of children’s school lunch box, and (ii) an index of intensity of plastic use in the household on a daily base. Similarly, in terms of disposal we define the following outcomes: (i) waste separation, and (ii) an index of frequency of plastic waste separation and disposal.

An intervention of this sort could lead to either weak or strong effects. A weak effect can be understood as a change in the individual’s knowledge/attitudes that may not necessarily translate into changes in behavior. In contrast, a strong effect will necessarily imply a change in behavior, which operates through actions and practices.

To account for the behavior of children and parents at home before and after the intervention, the experimental design also consists of the implementation of ex-ante and ex-post surveys, in combination with the environmental education program.This allows us to control for observable characteristics at the individual level, to understand households’ dynamics in relation with consumption and disposal of plastic, and to investigate to which extent households’ knowledge, attitudes and practices regarding marine plastic pollution problem and other environmental problems and environmentally friendly practices are affected in response to the program. Because we focus on direct and spillover effects of the program, these surveys are administered to both children and parents.

To account for the behavior of children and parents at home before and after the intervention, the experimental design also consists of the implementation of ex-ante and ex-post surveys, in combination with the environmental education program. This allows us to control for observable characteristics at the individual level, to understand households’ dynamics in relation with consumption and disposal of plastic, and to investigate to which extent households’ knowledge, attitudes and practices regarding marine plastic pollution problem and other environmental problems and environmentally friendly practices are affected in response to the program. Because we focus on direct and spillover effects of the program, these surveys are administered to both children and parents.

In the Chilean setting there is an extensive heterogeneity among schools, and these characteristics may affect the actual knowledge, attitudes and actions towards the use of plastics and the subsequent contribution to marine plastic pollution. We take these characteristics into account as part of the experimental design by generating a matched sample of schools based on the following observables: (i) coastal versus non-coastal towns, (ii) low- versus high-income schools, (iii) public versus private schools, and (iv) low- and high-level of environmental commitment of schools, based on their performance in the sustainable school program. This scheme allows us to generate a subset of schools that are comparable in the baseline. From our population of 105 schools, we randomly select 15 matched pairs of schools, being 15 schools assigned to the treatment group and 15 schools assigned to the control group, respectively

Schools

No

30 schools (i.e.,15 schools were assigned to treatment group, and 15 schools were assigned to control group).

Although the treatment is assigned at the school level, our unit of analysis the individual (i.e., pupils and parents/guardians). Based on official records, there are, on average, 2 courses per grade, and 25 students -on average- per class, so we expect to have a sample size of about 1320 pupils and 1320 parents/guardians,

15 schools control, and 15 schools treatment, respectively.

Because previous measures of knowledge and attitudes regarding plastic pollution problem are unknown, we focus on actions and practices. We use information from the Chilean National Survey of the Environment carried out in 2018. This survey only contains partial information of plastic disposal practices, however. These figures indicate that, in the Biobio region, 43.9% of individuals separate plastic recipients for recycling, with a standard deviation of 49.66%. Assuming a statistical power of 80% and without repetition, the minimum detectable effect for our total sample of 1,320 students is 7%.

Supplement 2 AEARCTR-0004650

Supplement AEARCTR-0004650

Reducing marine debris pollution by changing household behavior through children education