Public financial management (PFM) reform is a common part of many development initiatives. It generally involves promoting "good practices" in developing countries, embedded in frameworks like the Public Expenditure and Financial Accountability (PEFA) assessment (PEFA 2006). These include multi-year budgeting, competitive procurement, modern internal audit, and more. Such practices have proved effective in selected contexts and promise solutions to common problems in governments. Unfortunately, a growing literature shows that many governments do not solve their problems after years of adopting such solutions. Data reveal that governments commonly produce new laws that are not enforced and budgets that are not effectively executed, and suffer from weak capacities in distributed units (like line ministries and local governments) after many finished projects (Andrews 2006, 2011, 2013; Porter et al. 2011; Wescott 2009).
Studies tie these limits to a lack of realism in reform design and implementation (Andrews 2013; Andrews, Pritchett and Woolcock 2012; Booth 2011; Levy 2013; World Bank 2012). They argue, essentially, that reforms commonly fail to allow for necessary adaptation of external ideas to the realities in targeted contexts, often because the reform processes focus too narrowly on introducing the external good practice in principle and pay little attention to the practical difficulties of doing so in practice. Studies suggest, for instance, that such reforms pay insufficient attention to the political and administrative difficulties of effecting change, and that these difficulties commonly undermine reform results. Where studies see more effective and far reaching reform they often find that the externally nominated "good practices" are fitted to the targeted context through more adaptive processes that emphasize the real and practical issues of doing reform (like building reform support, testing and adjusting reform designs, and continually matching solutions and capacity realities and needs) (Andrews 2015; Andrews et al. 2014; Cabri 2014; Levy 2013).
These studies call for approaches that allow more realism in reform processes in developing countries, especially those supported by multilateral and bilateral donors. They do so knowing that similar calls have been made before; and that there are already many examples of such realism in the development community. There is, however, a challenge to identify these examples and describe what such processes look like in practice. This search leads quickly to a focus on bilateral donors. A small and interesting set of work suggests that such donors might have a comparative advantage in introducing more realism to PFM-type reforms in developing countries, given their own country’s recent experiences with doing such reforms. The argument is simply that development agencies from these countries can leverage the real experiences with doing reform in their own contexts when engaging with reformers in developing countries. They can, for instance, access experienced reformers to share lessons on issues like building demand for change, establishing political support for reform, and adapting reform ideas to context. Such lessons are often learned best through experience and remain tacit in those who have been through the experience. Bilateral agencies arguably have an advantage in accessing such people and their lessons, and can more effectively incorporate this valuable knowledge into their reform support than multilateral agencies.
This paper offers a novel analysis of this theory, asking whether Swedish development agencies working in the PFM field have leveraged the potential comparative advantage of the country’s own experience in supporting reform. The country’s own reforms have resulted in effective Sweden-specific adaptations of many of the good practices being promoted in developing countries today (including multi-year budgeting and modern accounting and audit). Academic descriptions of these reforms emphasize the processes by which they were adopted and the realism involved in such, and suggest the presence of many applied and tacit lessons one could see as valuable in developing countries (about testing reform ideas, for example, progressing gradually in reform processes, creating an urgent pressure for change, and building support for reform). The question asked here is whether Swedish development agencies bring these lessons into their support for PFM work, building on a potential advantage to promote realism in reform.
After introducing this question in an opening section, the paper provides a study of Swedish development agency engagement with PFM, in three domains: at the global level (where the development community has identified what "good" reforms should entail) and in two country-level experiences (Mozambique and Cambodia). The study uses process analysis to examine Swedish engagement in these domains (and reflects on "Swedish" involvement broadly, not on any
one development agency3). This approach offers a historical rendering of engagements since the late 1980s, based primarily on documentary evidence. There are limits to this kind of study, discussed in the methods section, but its strength is in allowing a view of reform support over time.
A conclusion notes that this view shows repeated attempts by Swedish development agencies to bring realism into their reform engagements. This has sometimes involved drawing on their own country's reformers and reform experience, although there does seem to be less of this than one might expect given the scope and success of the country’s own reforms. One explanation for this centers on evidence that Swedish reform engagements attempt to bring realism into their engagements by promoting a process-oriented way of doing development within its development agencies and with partner countries (not just by drawing on their own-country experience). This finding leads to a revision of the argument about how bilateral and multilateral agencies can promote realism in development. Another concluding observation notes that all efforts to bring realism have been less prevalent and effective in the past decade. Explanations are offered for this as well, including the growing importance of budget support in developing countries and the focus on highly specified and generalized PFM products—rather than process.