When is governance data good enough?
In the Eye of the Beholder: When is governance data "good enough"?
Takaaki Masaki, Tanya Sethi, and Samantha Custer. 2016. In the Eye of the Beholder: When is governance data “good enough”? Williamsburg, VA. AidData at the College of William & Mary and the Governance Data Alliance.
A growing number of governance data producers are investing significant time and resources to evaluate public sector performance in low- and middle-income countries. Yet, surprisingly little is known about how governance data is viewed by those it is intended to influence and whether the data we have today is “good enough” to usher in the policy change we are looking for. This report presents new evidence from a 2016 Governance Data Alliance (GDA) Snap Poll of public, private, and civil society leaders in 126 low- and middle-income countries to answer four critical questions:
- Delivery Channels: How do these leaders find or source governance data?
- Use: How is governance data used and for what purpose(s)?
- Influence: Which governance data do leaders find most useful – and why?
- Barriers: What are the most prevalent obstacles to the use of governance data?
Over 500 leaders shared their firsthand experiences in advancing reforms in their countries and the role of governance data in that process. Snap poll participants evaluated 29 governance data sources produced by a wide variety of multilateral organizations, bilateral agencies, and civil society groups.
Based upon their responses, we present four key takeaways.
- Broad-based communications still have sway, though the delivery channels leaders use to find governance data varies by where they work
- Governance data is predominantly used to conduct research and analysis; however, specific use cases appear to be shaped by different organizational mandates
- Most survey participants found governance data to be salient and helpful in their work, but this data is reportedly most useful when it is also perceived to be relevant and credible
- Governance data that fails to take into account the local context is seen as irrelevant and lacks credibility when it is not transparent in methods and assumptions.