Do African governments adopt transparency reforms just to make themselves look better? A response to Matt Andrews

Sep 18, 2013

This post was written by Paolo de Renzio, Senior Research Fellow at the Open Budget Initiative of the International Budget Partnership.

Matt Andrews from the Kennedy School at Harvard has just posted two interesting blogs (here and here) that use data from our Open Budget Index (OBI) to show that African countries tend to be more open about what they intend to do with public resources (transparency in budget formulation) than they are about what they have done with them (transparency in budget execution). This, Matt claims, could mean that countries adopt transparency measures “that make them look better but do not actually enhance functionality,” meaning that they do not lead to any real benefits.

Although we think that Matt’s arguments deserve a lot of attention, we do not think that the evidence clearly stacks up in their favor. Here’s why:

  1. It is not clear how Matt calculated countries’ “transparency gaps,” but our attempts to replicate his approach actually led to different results.
  2. The OBI methodology has in fact a built-in bias towards transparency in budget formulation, which might have influenced Matt’s analysis and findings.
  3. If we use an alternative method that simply looks at the public availability of budget documents, we do find that transparency in budget formulation is higher than transparency in budget execution, but the problem may not be as large as Matt finds, and could be due to various other reasons.
  4. A more useful approach to test Matt’s hypothesis would be to look at country trajectories over time, an area that deserves further research.


The issue that Matt raises is very important, but our preliminary analysis of OBI data does not fully support his arguments. In other words, there does not seem to be sufficient evidence that “countries generally focus on improving transparency in formulation to get better scores” on the OBI.

On a more general level, we agree with Matt that it is important to use international indicators like the OBI so that they promote change that is relevant and important to citizens’ lives and government performance in each country context, rather than promote cosmetic improvements to a government’s international image. We also think that there’s a need to recognize that the struggle for transparency needs to be put in a much broader context where a variety of domestic factors and dynamics shape governments’ willingness to open up their books, and the degree to which transparency can in turn contribute to positive development outcomes.

The road to “transparent transparency,” in other words, is still long and bumpy. We think that the OBI can be a useful tool in pushing governments and citizens along that road. Most of all, we think that the mere fact that it opens up opportunities to have this kind of discussion, based on detailed and disaggregated analysis of the data, can only be considered as a very positive and encouraging sign.

If you’re interested in our more detailed response to Matt’s blog, including in-depth analysis and examples, take a look at the extended version of this post.

One comment:

  1. It seems to me that budget preparation content is inherently more politically important and publicity worthy than budget execution. Maybe an analysis of all countries measures by OBI would find a pattern perhaps with a narrowing gap in more developed countries (although, maybe not!)

    I can remember a few years ago at a OBI presentation at an ICGFM conference where any delegates from Honduras questioned the index. It was well and dispassionately defended. Honduras improved their score in the subsequent test.

    But, is OBI a major driver for change? My sense is that PEFA is a bigger motivation because it is tied to donor funding. This could have a knock-on effect to improving OBI scores. That’s not to say that open budgets should not be a motivating factor for change, my sense is that it might be secondary to PEFA (and possibly OGP).

Leave a Reply

Your email address will not be published. Required fields are marked *