Syndicate content

At the fuzzy edge of public sector research

Philipp Krause's picture

There is something elusive about the workings of government. “We have virtually no adequate bureaucratic theory that can help us deal with the fundamental structural problems that exist with respect to the public sector. ” One might be forgiven for assuming this statement was written recently. Quite the contrary, this was E A Brett, expressing in 1986 what seems to be a perennial concern of public sector management.

The debate has shifted on a lot in the last 25 years, in particular with the 1997 World Development Report on the role of the state, and the research into public sector governance and institutions that followed it. But still there is a sense today that public sector research doesn’t quite deliver what’s needed by international development agencies and their government counterparts. More specifically, there is a mismatch between research into the “big picture” questions and the operationally relevant knowledge of how things are actually done.

The World Bank knows a lot about public sector management, and not just because Bank staff spend so much time working with governments. In the course of its engagement with governments around the world, the Bank produces a wealth of very rich empirical material, essentially case studies, many of which are never published or publicized. A Public Expenditure Review might, for instance, reveal in great detail why performance management was widely adopted in the health ministry of country X, even though it failed miserably at the ministry of transport.
The social science community struggles a bit to link the complexity of case studies to the large-n investigations that allow for more generalizable conclusions (never mind that this is also the topic of endless disputes over which approach is superior). For international development, this is of course literally not just an academic question. The generalized findings from public sector research needs to be relevant and helpful for reforming actual public sectors. Matt Andrews and others have pointed out how the current state of comparative research can sometimes even get in the way of appropriate advice to specific countries.

The World Bank’s new Public Sector Approach calls for a research agenda that combines the tacit, operational knowledge with formal research findings to deliver better results. The way forward, then, is to work on medium-range theories and hypotheses, addressed by appropriate methods to fill this gap. One interesting tool that could help is fuzzy-set qualitative comparative analysis (fsQCA).

Based on the work of Charles Ragin, fsQCA is a method that aims to understand why a certain phenomenon occurs, and works very well with a medium number of cases where different causal factors seem to be at work in a complex fashion. FsQCA assigns each case a membership score in the outcome (the dependent variable) and a number of causal conditions (the independent variables). Using Boolean algebra, one can then identify the particular combinations of causal factors that are sufficient to bring about the outcome. It is quite possible for different combinations to cause the outcome, thus making it conceivable for individual causal factors to have different effects depending on the overall combination of factors in that particular case. Free software is available to do the calculations, and there’s even a way to do fsQCA in STATA.

If one is interested in the relationship between, say, political institutions and fiscal outcomes in general, then fsQCA is quite limited. It is, however, just right for the conditions that brought about a certain policy change in a number of countries, over a certain period of time. In a recent paper, I use fsQCA to see why some countries adopt modern budgetary institutions that focus on outputs and performance, rather than keeping the old candle-end checks. Looking at 22 democracies of at least middle-income level, I find that countries generally respond to pressures to reform generated by having to deal repeatedly with fiscal crises. However, countries’ responses are not uniform and crucially, depend on the prevailing organizational culture. Only a combination of external pressures and the complex interaction between institutional arrangements and internal organizational culture can explain the variation between countries. The findings have very little claim to be relevant for countries which are not democratic, or of lower income. But there are many instances where an analyst would happily trade scope for depth when faced with a specific policy challenge.

A growing body of literature uses qualitative comparative analysis to study medium-range research questions about public sector management. For instance, Guy Peters uses this method to study the adoption of different administrative reforms in Western Europe and finds that policies diffuse unevenly between countries, and the strongest factor influencing diffusion is administrative culture, rather than fiscal crises or the political orientation of the government. For a comprehensive, up to date bibliography and other resources for fsQCA applications, the COMPASSS website is a good place to start.

A public sector research agenda will always rely on different styles of research and different methodological approaches. Qualitative comparative analysis is no panacea, but it might help to inform research into the kinds of questions public sector reformers are most interested in. It could also help to better leverage the empirical knowledge institutions like the Bank already have. 

Photocredit: flickr user n0ll

 

Comments

Submitted by Christian Schuster on
Philipp: I really enjoyed your post, not the least as it is the first one I recall in the GAC blog that is specifically devoted to discussing methodologies for undertaking public sector management research. In this regard, I had the impression that your very valid recommendation to strengthen the use of systematic (rather than ad hoc) qualitative comparative analysis, such as fuzzy-set analysis, could be of value not only to the Bank’s public sector management research, but to its research agenda overall (even though medium-range theories may, of course, be of particular use for public sector questions). As of now, the Bank seems to have undertaken extraordinarily few studies using fuzzy-set analysis in any of its research. In this regard and given that this research methodology has been developed quite some time ago and holds a lot of promise to derive operationally useful conclusions, I was wondering what your thoughts are on what has stood in the way of a more rapid adoption by the Bank and other development partners of fuzzy sets and other systematic QCA techniques?

Add new comment