The Contribution of Process Tracing to Theory-Based Evaluations of Complex Aid Instruments

Posted: 22 Aug 2016

See all articles by Johannes Schmitt

Johannes Schmitt

German Institute for Development Evaluation (DEval); University of Duisburg-Essen

Derek Beach

University of Aarhus

Date Written: October 1, 2015

Abstract

This article focuses on methodological challenges in evaluating complex program aid interventions like budget support. We show that recent innovations in process-tracing methodology can help solve the identified challenges and increase the strength of causal inference made when using case studies in demanding settings. For the specific task of evaluating the governance effectiveness of budget support interventions, we developed a more fine-grained causal mechanism for a subset of the comprehensive program theory of budget support. Moreover, based on the informal use of Bayesian logic, we have elaborated on how to increase the conclusiveness of empirical evidence for one part of the theorized causal mechanism. We argue that by establishing an explicit theorized mechanism prior to empirical research and by critically judging our evidence according to an informal Bayesian logic we can remedy some of the problems at hand in much case-study research and increase the inferential leverage in complex within-case evaluation studies.

Keywords: budget support, causal mechanisms, governance, methodology, process tracing, theory-based evaluation

Suggested Citation

Schmitt, Johannes and Beach, Derek, The Contribution of Process Tracing to Theory-Based Evaluations of Complex Aid Instruments (October 1, 2015). Available at SSRN: https://ssrn.com/abstract=2826466

Johannes Schmitt (Contact Author)

German Institute for Development Evaluation (DEval) ( email )

Fritz-Schäffer-Straße 26
Bonn, 53111
Germany

HOME PAGE: http://www.deval.org

University of Duisburg-Essen ( email )

Lotharstrasse 1
Duisburg, 47048
Germany

Derek Beach

University of Aarhus ( email )

Nordre Ringgade 1
DK-8000 Aarhus C, 8000
Denmark

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
277
PlumX Metrics