Blog

Designing ESID's research on Public Sector Reform

31 March 2014.
By Pablo Yanguas.
With “Effective States” as the first half of our name (and the entirety of our url!), ESID’s conceptual approach to the politics of development places a major emphasis on the ability of the public sector to deliver quality services. Most of our work in that regard has so far focused on the concept and measurement of state capacity: professors Kunal Sen, Antonio Savoia and Matthias vom Hau are three of our leading researchers in this area. However, sooner or later we would have to move beyond concepts and into processes, which is why we now have a dedicated project on public sector reform (PSR).
PSR is not a popular topic of research. It has been decades now since political science moved away from public administration, even if recent trends in democratisation, rule of law, political economy and conflict resolution analysis have all hinted at the centrality of the state institutions and organisations as a source of public authority. On the policy side, the development community has been dealing for some time now with the “failure” of the public sector reform agenda: beyond the walls of finance ministries and central banks where international financial experts tend to find intellectual allies, the reform of public administration and the control of public corruption have been constant headaches for donors and reformers alike. The politics of PSR are positively nightmarish.

When I began thinking about this project my social theory biases took centre stage: the original idea was to use basic theory (incentives, norms, cognitive inertia) as the analytical lens into the politics of public sector reform. However, peer reviewers interestingly pointed out that in so doing the project risked being disconnected from policy debates. As ESID’s ultimate goal is to improve development effectiveness, we decided to take on this concern and build it into the very core of the new project.
Much of the current “PSR zeitgeist” circles around guiding principles like Merilee Grindle’s “good enough governance” and David Booth’s “going with the grain“: the mistake of PSR was aiming to do too much, too quickly. A misplaced fixation with Weberian bureaucracy and an idealised version of Germany (or whatever other “good” state you prefer) has led practitioners to a permanent crisis of implementation: as Matt Andrews has argued, developing country states are more often than not engaged in isomorphic mimicry, replicating the forms of Weberian bureaucracy without actually reforming their day-to-day practice (for additional information on these arguments check out the ESID working paper that Badru Bukenya and I wrote).
In our new ESID project we want to explore to what extent the non-Weberian turn has yielded results in PSR practice. What we have done is build a comparative research project around the idea of competing new approaches to public sector reform: leadership, social accountability and policy adaptation, with Weberian administration still as the null hypothesis. But how do you test such vastly different approaches to reform?
First, you need to determine the unit of analysis: instead of analysing “The Public Sector” as a whole or any one service-delivery ministry subject to sector dynamics, we decided to look instead at core country systems or “PSR nodes” that most public sectors need in order to guarantee effectiveness over time – planning, civil service management, budget, auditing, anti-corruption and legislative oversight.
Second, you need data across time: comparative statics won’t do, as PSR processes usually take time. Our time frame would be determined by the reform agenda, which made us select the first decade of the 21st century – up until recent years – as the period during which PSR was deployed across the world in a somewhat coherent manner.
Third, you need data that is not publicly available in order to track organisational evolution. That means triangulation: interviews, internal documentation, and external assessments for each of our core country systems. In other words: plenty of fieldwork.
Fourth, you need to compare across contexts that are functionally equivalent, which means analysing the same PSR node in different countries or a range of organisations in the same country. We selected Rwanda, Ghana, Malawi and Uganda as a set of Sub-Saharan African countries which – at least in popular perception – embody all the levels of aggregate achievement in public sector reform, from outstanding success all the way to disappointing failure.
All told, the project encompasses six PSR systems in four countries, which means 24 cases, each of them comprising a decade of reform initiatives. It is not an experiment-level research design, but we feel that it is a big enough set of cases to be able to generalise with some degree of confidence. At the very least, we should be able to generate some interesting hypotheses and policy implications for the rest of the field.
It is also a major logistical endeavour, which is why the project relies on four field researchers, who will be coordinated from Manchester; all of us are young, early-career researchers, three of us are African, and oddly enough none of us are British. Together, we have roughly a year and a half to prove that public sector reform belongs as a major component of the politics of development. And it all begins in Cape Town a month from now.