What makes a successful behavioral insights (BI) unit?

The truth is, it differs in every context. Plus, there are many ways to build behavioral science into organizations. BIT knows this because we are the original “nudge unit” founded in 2010. But today, we do much more than nudge—including helping set up public and private sector BI units in Chile, Argentina, the UAE, and the US.

In this work, we’ve noticed a few common factors to success. Behavioral science knowledge and the motivation to continuously learn and improve are key. 

Ceibal, the Uruguayan government’s digital technology center for education innovation, embodies these attributes and more. In early 2023, with support from the Inter-American Development Bank, they reached out to BIT for help structuring their own BI unit, the Behavioral Insights/Science Lab.

How we helped the Lab think about its future

Before our work together began, Ceibal’s Lab had already made an impact. They developed two interventions, one that boosted teacher attendance to online summer school workshops and another that led to more female students getting accepted to a computer science program.

To build on this progress, BIT supported the Lab’s consolidation strategy. Ceibal wanted us to analyze their current state, identify potential barriers they may face in development, and think about opportunities for the future. APPLES, a framework BIT designed to help in the creation of public sector BI units, guided this work.

First, the BIT team interviewed stakeholders critical to the success of the Lab, including public officials from the Ministry of Education, the National Administration for Public Education (ANEP), and Ceibal team members. 

Ol Beun, BIT Principal Advisor, leading a TESTS workshop with the Ceibal team.

Then, we held workshops at Ceibal to develop a theory of change, including short- and medium-term results the Lab wanted to achieve, and the activities to help get there.

BIT and Ceibal also collaborated online in diagnosis workshops to identify resources the team would need to grow, barriers to progress, and opportunities to address them.

Ceibal and BIT team members.

Pitfalls and how to overcome them

We identified a few barriers, including:

Barrier #1—Risk of irrelevance: In an effort to prove its value, some BI units solely focus on small-scale and short-term interventions in areas the team is familiar with in order to achieve rapid results. As a consequence, they avoid working on new priority topics or projects with broader exposure where there is potential for impact, but it feels less certain. 

• How to overcome it: We recommend that the Lab builds a diverse portfolio of work—from quick wins to high priority projects—focusing on topics that will advance outcomes that the government cares about most. Presenting behavioral science work in a way that is easy for non-experts to understand and see the value of is another tactic to overcome this barrier. Accessible formats could look like short presentations, reports, and publishing findings externally.

Barrier #2—Difficulty scaling results: While the Lab had already completed successful behavior change projects before our work together began, they found it difficult to ensure the evidence was scaled.

• How to overcome it: Consider scaling at the very beginning of every project. John List’s “Voltage Effect” offers practical factors to take into account, such as the risk of finding false positives, ensuring that the intervention will represent the target population, avoiding components that cannot be scaled, and balancing the cost-benefit of the intervention at scale. These prompts will guide BI units toward efforts that, if successful, will produce scalable outcomes.

Successes in student attendance

Ceibal put these recommendations into action right away. Student absenteeism is a major political priority in Uruguay. The Lab and BIT evaluated a multi-component communications intervention that led to a 6% reduction in primary school absences. Learn more about this work here.

While designing the intervention, we accounted for every factor of the “Voltage Effect” as well. For example, to avoid the risk of false positives, we used past behavioral science evidence and replicated personalized letters to parents—an intervention that had worked in other contexts.

As a result, Ceibal and ANEP are scaling personalized letters to parents of 1st to 3rd grade students nationwide—nearly 50,000 students total. If successful, it could lead to 110,000 days of school attended that would be missed otherwise.

Let’s enhance your organization with behavioral science

Ceibal’s Behavioral Insights/Science Lab is setting a new standard for evidence-based policymaking in Uruguay. The Lab’s four-person team works flexibly to identify behavioral projects that will advance policy priorities, and rigorously evaluates them to learn what works. This will, all together, make better use of public resources over time.

We hope that other governments and organizations across Latin America find inspiration in Ceibal and explore embedding behavioral science into their own work. We would be happy to talk through your ideas. Contact BIT here if you’d like to discuss enhancing your organization with behavioral science.