Fostering Trust to Save Lives: Evidence from Organ Donation in Tunisia: Guest post by Christina Sarah Hauser

This is the 7th in this year’s series of posts by PhD students on the job market.

Every year, thousands of people die waiting for an organ transplant. The main constraint is not technology or medical expertise: in many low- and middle-income countries (LMICs), organ transplantation is technically feasible, however, the supply of donated organs is far below the need. Why? In Tunisia, our scoping research found that people often lack basic knowledge about how organ donation works, mistrust the institutions in charge, and worry about trafficking and religious prohibitions - concerns which are unfounded as Islam strongly encourages organ donation.

Policy changes may not always be the right answer. In 1991, Tunisia introduced “presumed consent”, following the example of several European countries, which saw large increases in organ donation after making being a donor the default option.  But presumed consent only works in societies with high institutional trust: in Tunisia, officials anticipated public backlash and, in practice, always seek family consent when the deceased’s preferences are unclear. Unfortunately, family refusal rates remain high, leaving the country trapped in an inefficient equilibrium where demand far exceeds supply

Given the low levels of awareness and trust, we hypothesized that local experts would be in the best place to spread awareness of organ donation and foster trust in medical institutions. My job market paper asks whether an expert-led informational intervention can change young adults’ minds about organ donation. More importantly, can it actually change behavior, leading to an increase in organ donor sign-ups?

The intervention

Together with Tunisia’s National Center for the Promotion of Organ Transplantation, we designed an informational intervention and evaluated it within a randomized controlled trial (RCT) at a Tunisian university, enrolling 980 undergraduate students in spring 2025. The intervention was randomized at the classroom level, took about 30 minutes in total, and had three core elements:

1. A patient testimonial: a short video of a heart transplant survivor from rural Béja.
2. An expert-led informational session: medical doctors and technicians came to the classrooms to explain the medical procedure and legal framework in Tunisia, as well as religious rulings in favor of organ donation.
3. A Q&A session: participants could voice doubts and ask questions to the experts.

We collected a detailed baseline prior to the intervention, including measures of students’ social network to test for spillovers. The baseline data confirms that randomization was successful. The main challenge we faced was measuring real behavior: in Tunisia, adults can formally declare their wish to be an organ donor by adding the word “donor” to their national ID but there is no publicly accessible donor registry. Thanks to a collaboration with the Tunisian Technical and Scientific Police, we were able to offer - and measure - ID changes on campus for two days after the intervention.

What are the main findings from the experiment?

Information works:

●        Donor registration increased. 11 students from the treatment group signed up, compared to only 3 students in the control group, which translates to a 1.8 percentage point increase (Figure 1, Panel A). This means that sign-ups among the treated more than doubled compared to the control group.   

●        Improved knowledge and higher institutional trust emerge as the most important mechanisms: We use a detailed endline survey to assess potential mechanisms, aggregating different items to indices as specified in our pre-analysis plan. Treated students scored 1.1 and 0.8 standard deviations higher on medical and legal knowledge. Institutional trust indeed increased by 0.7 standard deviations (Figure 1, Panel B).

●        Religious misperceptions decreased: Participants became more likely to know that Islam encourages donation. The “religious beliefs” mechanism is less robust though and does not survive corrections for multiple hypothesis testing (Figure 1, Panel B).

Panel A. Estimated treatment effects on ID changes
by expected family approval at baseline and for the whole sample).

Panel B. Estimated treatment effects on the secondary outcomes aggregated to indices

Notes: ***  and * indicate statistically significance at the 1% and 10% levels, respectively

Families matter

One of the most striking findings is the role of family. For heterogeneity analysis, we created an index of expected family approval based on students’ expectations at baseline, parental education and profession (parents who work in medical professions are more likely to support organ donation). Students whose families are likely to approve of organ donation were 7.4 percentage points more likely to change their IDs after the intervention. For those whose families disapproved, the treatment had virtually no effect.

While we should be cautious about drawing causal conclusions based on this finding, this pattern aligns with qualitative evidence suggesting that organ donation is not seen as an individual decision in Tunisia—it’s a family decision. Our further heterogeneity results support this interpretation: treatment effects on donor registration are driven primarily by men, who may face fewer constraints from family attitudes than young women. To be effective, policies aimed at increasing organ donation must therefore recognize and engage with the family dynamics that shape young adults’ decisions.

The power of peers

Our network data revealed another interesting pattern: students who attended the session alongside their friends were more likely to register as organ donors. However, we found no significant spillovers to students who did not attend. This suggests that merely knowing someone who attended the informational session is not enough - the role of the expert conveying correct information seems crucial. Still, the shared experience of attending the information session with friends seems to amplify the effect, highlighting the importance of social context in shaping behavioral change.

Policy implications

The intervention is cost-efficient, especially compared to alternative treatments, such as kidney dialysis, which - if medically feasible - are very costly for health systems and painful for the patients. What else can policymakers take away from our study?

1. Trusted voices matter: Interventions led by medical experts can build institutional trust, with a direct impact on behavior.
2. Keeping family dynamics in mind is key: In non-Western contexts, young adults may require their parents’ approval to sign up as deceased organ donors. Extending outreach to parents and older generations could therefore increase treatment impact. This finding is likely to apply to other decisions beyond organ donation.
3. Leverage peer dynamics whenever possible: The students who attended the intervention together with their friends were more likely to become organ donors. This suggests that structuring interventions so that people attend with their social support group can amplify effects.

Key takeaways and future work

Organ transplantation saves lives but only if enough people become donors. Our study suggests that expert-led, context-sensitive interventions can make a difference even when general levels of awareness and institutional trust are low. The proposed intervention is cost-efficient, making it a realistic policy option for many LMICs.

Our findings come from young adults at a single Tunisian university. Future work should also test whether scaling this approach beyond universities is feasible. Whether effects on knowledge and trust persist long-term remains an open question. A follow-up survey later this year will shed light on persistence, and we have scheduled a second round of the trial - so stay tuned!

Christina Sarah Hauser is a Postdoctoral Researcher at Collegio Carlo Alberto in Turin, Italy. She received her PhD in Economics from the European University Institute in 2025.

This research was only possible thanks to the intellectual and practical support of the Tunisian Center for the Promotion of Organ Transplantation (CNPTO). We would like to express our gratitude to the CNPTO and especially to Dr. Boutheina Zannad, Dr. Inès Jaafar, Khadouja Ben Nakissa and Mejda Chaouachi. We also thank Leila Triki, Sana Mami, Imen Bouhestine, Lamia Ben Fdhila, Emnaa Chahed, and the faculty of the South Mediterranean University