Strengthening Evidence-Informed Policymaking: A Randomized Experiment on Capacity Building and Evidence Intermediation in Ethiopia

How to cite

Abate, Gashaw T; Abay, Kibrom A; Hill, Ruth; Nigus, Halefom; Oliver, Kathryn. 2026. "Strengthening Evidence-Informed Policymaking: A Randomized Experiment on Capacity Building and Evidence Intermediation in Ethiopia." AEA RCT Registry. AEARCTR-0018969. https://doi.org/10.1257/rct.18969-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Gashaw T Abate* , Kibrom A Abay , Ruth Hill , Halefom Nigus , Kathryn Oliver

* Primary investigator

Registry ID
AEARCTR-0018969
DOI
https://doi.org/10.1257/rct.18969-1.0
First registered
Last updated
Status
In development
Trial dates
2026-08-15 – 2027-12-31
Countries
Ethiopia ()
Registry record
http://www.socialscienceregistry.org/trials/18969

Abstract

Despite substantial investments in research capacity across Sub-Saharan Africa, research-based evidence rarely shapes policy decisions, due to both demand- and supply-side constraints. This study tests two interventions among high-level Ethiopian policymakers to relax these constraints. The first is a capacity-building and training program – delivered through the African Leadership Excellence Academy (AFLEX) – targeting policymakers' awareness, interpretive skills, and trust in research evidence. The second combines this training with an on-demand evidence synthesis and intermediation service, housed within the Ethiopian Policy Studies Institute (PSI), which provides AI-assisted rapid syntheses of research-based evidence relevant to specific policy questions. Using an experimental design, the study will estimate the independent and combined effects of these interventions on policymakers' evidence awareness, capacity, and use. A systematic baseline assessment of individual, institutional, and political economy factors shaping evidence uptake complements the experimental arms, enabling analysis of how organizational and contextual conditions mediate intervention effects. By embedding both interventions in existing institutional structures, the study prioritizes sustainability and generates transferable lessons for strengthening evidence ecosystems in Ethiopia and comparable low- and middle-income settings.

Intervention

The experiment has two interventions.

Intervention 1: Capacity Building and Training (T1)

T1 is a structured training program on research-based evidence for policy decision-making, implemented by the African Leadership Excellence Academy (AFLEX). It targets directors at ministries and their accountable agencies, commissions, and institutes engaged in economic policy across key sectors, including agriculture, industry, trade, energy, transport, and labor and skills. The program addresses both demand-side constraints – such as limited awareness of evidence-informed policymaking, inadequate skills to access and interpret research, and weak trust in scientific evidence – and supply-side constraints related to limited investment in policy-relevant research. Training combines theoretical and practical modules that draw on real-world cases to illustrate the value of research evidence in policymaking, while building participants' capacity to commission, interpret, and apply evidence in their work. Sessions are facilitated by experts with demonstrated experience in evidence production, intermediation, and use.

Intervention 2: Capacity Building, Training, plus Evidence Intermediation (T2)

T2 combines the T1 training with evidence translation and intermediation service hosted at the Ethiopian Policy Studies Institute (PSI). The service responds to policymakers' requests by delivering rapid, systematic syntheses of existing research evidence on specific policy topics, communicated in accessible, targeted formats. Synthesis production is supported by AI tools to enhance efficiency and scale, with policymakers involved in co-creation where feasible.

T1 primarily addresses demand-side barriers through training alone; T2 tests whether pairing training with institutionalized, supply-side evidence intermediation support generates additional gains in evidence uptake and use.

Experimental design

The study uses an individual-level randomized controlled trial (RCT) in which approximately 360 senior policymakers are randomly assigned, stratified by ministry, to one of three arms:

(i) Control (C): no intervention;
(ii) Capacity Building (T1): training on research evidence awareness, interpretation, and use; and
(iii) Capacity Building + Evidence Synthesis (T2): T1 training and access to evidence synthesis service with opportunities for co-creation.

This design supports three causal comparisons – T1 vs. C (effect of training alone), T2 vs. C (combined effect of training and evidence supply), and T1 vs. T2 (marginal effect of the evidence synthesis service conditional on training) – enabling identification of both demand- and supply-side barriers to evidence use in policymaking.

Design details

Randomization method
Randomization will be done in office by a computer.
Randomization unit
Individual policymaker
Clusters
0
Observations
360
Arms
120 policymakers’ control, 120 policymakers receive training on evidence-based policymaking, 120 policymakers received training plus evidence synthesis
Primary outcomes
Perceptions and attitudes: Policymakers' self-reported awareness, capacity, and trust regarding the value and use of research-based evidence in policy decision-making. Behavioral demand for evidence: Stated and revealed demand for research evidence, captured through the number of evidence requests submitted, researcher engagement, use of evidence repositories, and reference to evidence in policy processes. Organizational procedures and processes: Extent to which policymakers establish or strengthen institutional procedures – including evidence review processes, policy option evaluation frameworks, and research partnerships – to support evidence-informed decision-making. Policy outcomes: Development of new policies, legislation, or strategies, and concrete revisions to existing policy documents or resource allocation decisions, that are demonstrably informed by research evidence.
Secondary outcomes
Preferences and willingness to pay: Policymakers revealed preferences for specific evidence attributes (accessibility, timeliness, source) and their incentive-compatible willingness to pay for research evidence.

Cited by

No citation data available for this registration yet.

Keywords