Evidence for the information-accountability-school performance link

After months and months of proposal writing, revision, and planning, I was excited to learn over the past few days that Princeton approved the human subjects protocol and all the other bureaucratic details for an evaluation of the Uwezo initiative in East Africa. I described this project briefly in my initial post — but the main idea is that Uwezo will try to improve education outcomes in Kenya, Tanzania, and Uganda by providing various forms of information to citizens. First, they go to randomly selected households and administer a series of tests to children; second, they provide immediate results to the parents; third, they provide printed information for parents and for the school and community more generally about what they can do to improve education directly (e.g., read to kids) and through policy change and oversight; and fourth, they reinforce that information and put it into broader perspective through information campaigns that describe the results of the assessment. There are some other nuances that I’ll write about later, but that’s the gist.

In our work on this project, Dan Posner, Lily Tsai, and I will be trying to assess the degree to which information empowers citizens to effect meaningful changes with concrete benefits for the education of children. Much of the research will consider whether the information provided is actually “news” to parents and other citizens; whether such information changes people’s views on the services they are receiving; and whether it effects the likelihood of their taking any action. We’ll get started with extensive field research in Kenya this June.

But in the meantime, it’s worth discussing the highly relevant, recent World Bank study, Making Schools Work: New Evidence on Accountability Reforms.

World Bank, Making Schools Work

Education – Making Schools Work.

This is an excellent review of a series of recent studies that have sought to improve education outcomes particularly in developing countries. What’s particularly important about the study is that starts with the premise that there is a huge amount of variation that is NOT due to differences in resources allocated. We certainly know in the United States that money doesn’t solve all of the problems of providing acceptable universal education.

They focus on a series of studies that involved randomized interventions in various  countries, including efforts to effect teacher performance, and school management. But most relevant for our study is chapter two, which investigates a bunch of studies — conducted in India, Pakistan, and Liberia. So far, the results have been pretty mixed, and not terribly encouraging in the sense that information campaigns have not had overwhelming effects on citizen activity or educational outcomes. For example, Banerjee et al (2009) conducted a study in the state of Uttar Pradesh in India, and found that the provision of information about local governance and self assessment tools had only a slight effect on citizen awareness, and virtually none of participation, or learning outcomes, when compared with households from control villages. On the other hand, only a few studies have been conducted, they differ in several ways, and as the authors of the World Bank study conclude, this type of research is in its infancy, and we still know very little about whether and how such interventions can have a measurable effect. To our knowledge, no similar study has been conducted in East Africa. The idea that better information should help citizens to realize the promise of democratic governance is too important to discard based on just a few studies, and while we will let the results of our study speak for themselves, it will be extremely important to understand if and how new information influences citizen participation in development.

Leave a Reply