Key message 4 – The way information is provided matters.

There are some key elements to pay attention to when designing an information intervention.

  • Making information stand out: Information has to be provided in a way that captures people’s attention. Faced with complex problems and limited attention spans, people often ignore valuable information, and benefit from summaries highlighting key relationships (Hanna et al. 2013). Information interventions may thus fail to improve outcomes if the information they provide is not absorbed by programme participants.
  • Certification is useful: Information is not always trusted. For example, in a recent trial in South Africa, providing jobseekers with certified skills assessments substantially improved employment and earnings, while giving jobseekers the same information without a formal certificate did not 8(Carranza et al. 2018). This suggests that even if jobseekers are aware of their skills, they may not be able to credibly signal these skills to potential employers.
  • People might ignore the new information if it threatens their self-esteem: There is growing evidence that people interpret information in ways that help them maintain a positive image of themselves. For example, Mobius et al. (2011) use performance on an IQ test to show that people will change their self-assessment too much when told they likely did well, and too little when told they likely did not do well. The same subjects do much better at this “belief updating” when their self-esteem is not at stake. In a labour market context, providing information about one’s skills or job prospects is likely to generate a similar type of behaviour. Hence information provision is likely to work better if it is framed in terms of “good news” that does not threaten jobseekers’ self-image. Affirming the jobseekers’ self-worth (such as giving them positive information on other skills) can also help accurate processing (Cohen et al. 2000).

A woman looks at a poster advertising jobs on the street, Addis Ababa, Ethiopia. (Photo by In Pictures Ltd./Corbis via Getty Images)


  • 8 Effects in the ‘private’ treatment (without the certificate) were always smaller than the ‘public’ treatment (with certificate), and often statistically so.