How can we encourage citizens to comply with desired civic behaviors? In a CDDRL seminar series talk, Salma Mousa, assistant professor of Political Science at UCLA and former CDDRL postdoctoral fellow, explored this broader question via a field experiment in Lebanon. In conjunction with a municipality and local NGO, Mousa and her team evaluated the effectiveness of a waste sorting intervention.
In 2015, some of Lebanon’s primary landfills reached capacity, forcing displaced waste into the streets and prompting public outcry. Lebanon's crisis is not for lack of money; the country spends ten times more than nearby Tunisia despite having only half the population of Tunisia. This suggests that Lebanon’s issue reflects mismanagement rather than a lack of resources.
A key component of this mismanagement is a lack of sorting at the source of waste. Effective sorting, Mousa argues, requires collaboration between citizens, civil society, and government. Overcoming this collective action problem does not just require physical infrastructure and intrinsic motivation; it also requires that people trust that their neighbors and government will do their part.
To test their sorting intervention, Mousa and her collaborators chose the small, wealthy, and predominantly Christian town of Bikfaya. The town is characterized by high levels of social cohesion and a “green” reputation that is central to its identity.
Working with the municipality and an NGO called “Nadeera,” the team divided the town into neighborhoods, randomly assigning treatment and control. The treatment group received a box with QR codes they could put on their trash bags and an app where they could access feedback on their sorting. They were given instructions on proper waste management and told to sort their waste into recycling, organic materials and other — sticking their personal QR codes on each bag.
After pickup, inspectors at the nearby waste management facility would use the app to provide personalized feedback on sorting quality, giving participants the opportunity to improve.
This intervention makes trash sorting a sanctionable behavior, with social pressure to enforce it, because participation is visible to neighbors via the QR code stickers placed on their trash bags.
The team examined three distinct outcomes. First, the quality of sorting. Second, participation in a raffle for “green” prizes, designed to measure the impact of the intervention on other climate-friendly behaviors. Finally, they measured participation in volunteer opportunities for environmental initiatives.
Two months after the intervention, the program improved sorting quality by an average of 14 percent. That said, at the twelve-month mark, the effect was null. Eight months in, the program and app feedback ceased, making it difficult to distinguish between diminishing long-term effects and lack of sanctioning.
Treated participants entered the raffle at two times the rate of the control group, but the mechanisms behind this increase remain unclear. The rise in uptake could be attributed to behavioral change or familiarity with the NGO as a result of treatment.
On the volunteering measure, the treated group saw a 7% negative effect, meaning they were less likely to sign up for local environmental initiatives if assigned to treatment. Mousa and her collaborators theorize that this is due to moral licensing, or the feeling that they have already done their part.
While the effects of the primary outcome became null after a year, the treated group did see a substantial improvement in sorting quality — a big win for the town on environmental and economic measures. Future iterations of this intervention will include consistent monitoring or cash benefits to promote prolonged participation.
