- Successful reforms have started by identifying and framing the problem
- Preparation and political economy analysis can help identify windows of opportunity
- Broad engagement with leaders and networks can help to build constituencies for change
Recent research findings suggest that the process of reform (how a problem is identified and a solution developed) is more important than its content (World Bank, 2012, p. 10). A growing body of qualitative research explores process approaches, but does not form a systematic and rigorous evidence base. In a review of evaluations of process approaches, Walton (2011) finds mixed outcomes on development effectiveness. Two recent sets of research (Andrews, 2013; Tavakoli et al., 2013) provide frameworks designed to be adapted to context-specific diagnostics and evidence (Harris & Wild, 2013, p. 4).
Andrews’ (2013) wide-ranging analysis concludes that what has worked is what he calls ‘problem-driven iterative adaptation’ (PDIA). This involves step-by-step flexible experimentation with relevant solutions (‘purposive muddling’ and ‘problem-driven learning’). It draws on (a broad scanning of) external and internal ideas and engagement, with many agents (not solitary champions) playing multiple roles to form hybrid solutions. Andrews emphasises that ensuring reforms are politically and practically feasible requires patience (to allow iterative learning and change) and a focus on small next steps, not final solutions.
Features of problem-driven iterative adaptation
|Issue||Reform as signals||Problem-driven iterative adaptation|
|Context||Context is ignored, or only visible context is considered||Contextual complexities are revealed and addressed through problem-driven learning|
|Content||Some (visible) elements of externally-defined best practice are copied – ‘isomorphic mimicry’||Content is found and fitted through purposive muddling involving ongoing experiential learning and feedback|
|Agents||Narrow sets of high-level agents are relied on to champion reform; implementation by distributed agents is assumed to happen ‘by edict’||Broad sets of agents are mobilised into communities of change by conveners, connectors and motivators, fostering the emergence and diffusion of local solutions|
|Result||New forms do not improve functionality||Slow, incremental process whereby localised hybrids emerge to solve pressing local problems; new forms may not look impressive but they work|
|Source: Andrews, 2013, p. 216|
Tavakoli et al. (2013) examine four case studies of aid programmes that addressed governance constraints in service delivery. They conclude that how programmes are designed and implemented affects whether they are able to gain domestic traction and support institutional change. They identify six characteristics that support institutional change and improved outcomes, and give examples of current and proposed practice for each of these enabling factors. Some of these may already be standard practice for some donors.
Enabling factors in aid design and delivery
|Enabling factor||Current practice (stylised)||Proposed practice|
|Windows of opportunity||Donor country strategies based on needs assessments||Weigh need against opportunity to effect change|
|Tangible political payoffs||Donors claim political neutrality||Accept that aid is inherently political and work with the political incentive structure|
|Building on what’s there||Tendency to want to start afresh with a revised legal, policy or regulatory framework||Implement existing framework, however imperfectly, then adjust|
|Moving beyond policy advice||Technical assistance engaged to advise on the content of policies||Technical assistance engaged to support implementation of policies|
|Acting as facilitators||External agents provide technical advice for governments to ‘take or leave’||External agents help to facilitate and mediate a local dialogue about problems and solutions|
|Adaptive and responsive to lessons learnt||Pre-defined logical frameworks that lock donors and implementers into a set of activities||Flexible frameworks that do not use pre-defined targets to judge performance|
|Source: Tavakoli et al., 2013|
Identifying the problem
A number of experts (Andrews, 2013; Fritz et al., 2009) stress that the reform process has worked best when it has started with problem identification and framing. Fritz et al. (2009) find that such problem-driven analysis can promote innovation and support for proactive campaigns of change.
One example given by Andrews (2013, pp. 136-137) is the success of support to Nepal’s Health Sector Programme 2005-2010. Despite political upheaval and regime transfer, successive governments remained committed and implementation stayed on track. He concludes this was partly because the intervention focused on problems that broad groups cared about deeply (providing basic health services to underserved people). These problems were framed using data about outcomes and outputs that politicians and bureaucrats were sensitive to (e.g. how many children were dying of measles because they had not been immunised, and how many hospitals were functioning in rural areas).
Windows of opportunity
Various strands of institutional theory posit different explanations for how institutional change occurs – either incrementally or through sudden exogenous shocks. In the incremental perspective, institutions are viewed as objects of political contestation, and change is a slow process of adaptation by various actors (North, 1990, p. 18). In contrast, a structural analysis suggests that institutions tend to resist change and only moments of exogenous shock or major power struggles between elites can shift them (Leftwich & Sen 2010). Public sector institutions can be initiators of change, but also sources of legitimation, implementation and enforcement of new institutions.
Leftwich and Sen (2010, p. 40) and Leftwich and Wheeler (2011, p. 9-10) note that institutional change often has critical junctures, ‘triggers’ and windows of opportunity. Understanding these openings – ‘being ready’ for ‘planned opportunism’ (Eyben, 2010, p. 11) and identifying what is feasible – is a key skill for local leaders and donors. Yet ‘path dependency’ and ‘institutional stickiness’ often limit institutional reform or innovation (Leftwich & Sen, 2010, p. 40). Bunse and Fritz (2012) identify the following political economy factors:
Political economy factors for public sector reform
|Grouping of +/- factors||Political economy factor||Frequency across countries|
|Factors that can trigger or induce initiation of public sector reform (PSR)||Fiscal pressures or crises||Some|
|Global economic integration||Many|
|Donor pressures and involvement||Some|
|Disincentives for implementing PSR||Time horizon: disincentive for current government to incur cost if benefit of PSRs cannot be reaped during tenure||Many|
|Bureaucratic resistance increasing the political cost of implementation||Many|
|Rent-seeking related to maintaining privileged insiders and clientelism||Many|
|Limited demand for public sector reforms||Many|
|Proportional electoral rules||Some|
|Conditions strongly or possibly favouring PSR adoption and implementation||Rapid and sustained growth||Few|
|Post-crisis rebuilding efforts||Few|
|Government seeking legitimacy||Many|
|Legacies of meritocratic public sector||Some|
|Majoritarian electoral system||Some|
|Significant demand for public sector reforms||Few|
|Proximate factors and strategies that may facilitate PSR implementation||Policy entrepreneurs||Some|
|Quick win strategy||–|
|Source: Bunse and Fritz, 2012, pp.21-22|
Tavakoli et al. (2013) highlight practical approaches for identifying windows of opportunity. Identifying opportunities is pertinent throughout the intervention cycle – from being in the right place to spot opportunities before a reform programme starts (case 1), to understanding the political economy and possibilities for reform during programme design (case 2), to keeping abreast of changing contexts and relationships (case 3). Other evaluations have highlighted the importance of being in the right place with committed support: Folscher et al. (2012) find that for public management reform in Malawi, even though political/institutional drive for reform was absent, donor inputs successfully introduced and sustained reforms until conditions were right for their acceleration.
|Windows of opportunity|
|Case 1: Investment in programme design – local government development programmes, Uganda||A large upfront investment in a year-long programme design and consultation phase revealed entry points. The design process included understanding the situation in each district, developing a mechanism to incentivise performance, and building understanding and support for the new grant mechanisms.|
|Case 2: Governance assessment – rural water and accountability programme, Tanzania||A ‘strategic governance and corruption assessment’ (including a power and change analysis) informs programme design. Updated every two years to help ensure entry points remain relevant and identify where adjustments are needed.|
|Case 3: Trial and error – support to the Strategy and Policy Unit (SPU), Sierra Leone||The Africa Governance Initiative (AGI) moved advisors to different ministries as it became apparent where there was true reform drive.|
|Source: Tavakoli et al., 2013, p. 36|
- Discussion at the ODI seminar on Unblocking results: can aid get public services flowing? in 2013 highlighted common themes emerging from recent research (Andrews, 2013; Booth, 2012; Tavakoli et al, 2013) and older research (Easterly, 2007; Centre for The Future State, 2011).
- Recommendations for donors to adopt a flexible process approach have been in circulation since the 1980s (for example, see Brinkerhoff and Ingle, 1989, on structured flexibility), if with little traction on programming (see Walton, 2011). Features of this type of approach have appeared in work on organisational learning and in recent theory on complexity and development (Ramalingham & Jones, 2008; Jones, 2011b).
- Only a small number of programmes employ this terminology and an even smaller number of studies evaluate their effectiveness. Most of these studies focus on success stories to support an argument about the usefulness of the process approach. Difficulties in developing the evidence base include: the complexity of institutional reform; inconsistent and infrequent use of process-related terminology; and lack of explanation of programmes’ implementation processes in evaluations (Walton, 2011).
- This ‘science of muddling through’ has antecedents in earlier organisational sociology (e.g. Lindblom, 1959).
- Andrews, M. (2013). The limits of institutional reform in development. New York: Cambridge University Press. See document online
- Bunse, S., & Fritz, V. (2012). Making public sector reforms work: political and economic contexts, incentives, and strategies (Policy Research Working Paper No. WPS 6174). Washington D.C.: World Bank. See document online
- Eyben, R. (2010). Hiding relations: The irony of ‘effective aid’. European Journal of Development Research, 22(3), 382-397. See document online
- Fölscher, A., Mkandawire, A., & Faragher, R. (2012). Evaluation of public financial management reform in Malawi 2001–2010 (Final Country Case Study Report). Stockholm: Sida. See document online
- Fritz, V., Kaiser, K., & Levy, B. (2009). Problem driven governance and political economy analysis. Washington, D.C.: World Bank. See document online
- Harris, D., & Wild, L. (2013). Finding solutions: making sense of the politics of service delivery. London: Overseas Development Institute. See document online
- Leftwich, A. & Sen, K. (2010). Beyond institutions: Institutions and organisations in the politics and economics of poverty reduction – a thematic synthesis of research evidence. DFID-funded Research Programme Consortium on Improving Institutions for Pro-Poor Growth (IPPG), September 2010. University of Manchester. See document online
- Leftwich, A., & Wheeler, C. (2011). Politics, leadership and coalitions in development: Findings, insights and guidance. Developmental Leadership Programme. See document online
- North, D. (1990). Institutions, institutional change, and economic performance. New York: Cambridge University Press. See document online
- Tavakoli, H., Simson, R., Tilley, H., & Booth, D. (2013). Unblocking results: Using aid to address governance constraints in public service delivery. London: Overseas Development Institute. See document online
- Walton, O. (2011). Evidence for the effectiveness of a process approach (GSDRC Helpdesk Research Report 755). Birmingham: GSDRC. See document online
- World Bank. (2012). The World Bank’s approach to public sector management 2011-2020: ‘Better results from public sector institutions’. Public Sector & Governance Board, Poverty Reduction and Economic Management. See document online