Closing a Gap? The Role of Bottom-up Costing Within MTEFs


Posted by Michael Di Francesco[1]

The medium-term expenditure framework (MTEF) is one of the defining planning and budgeting practices of contemporary public financial management (PFM).[2] The PFM literature is replete with analysis of the relative strengths and weaknesses of different types of MTEF, but all are generally characterized by two features. First is the use of multi-year estimates of expenditure to account for the consequences of current year budget decisions; in other words, an MTEF is a critical ‘framing’ mechanism for more strategic budget preparation. The second feature is the emphasis placed on an iterative budget process that aims to reconcile the disciplinary function of ‘top-down’ expenditure envelopes set by finance ministries, and rigorous ‘bottom-up’ costing of programs by spending ministries.

The role of ‘bottom-up’ costing, crucial as it is to the accuracy of baseline funding estimates and the relative assessment of spending programs, has received very modest attention in the literature. Within the context of the MTEF, costing is reliant on three elements: the forecasting capacity of finance ministries, the quality of program-specific information that usually resides with spending ministries, and the reasonableness and consistency of assumptions, especially about behavioral responses to program interventions. Whilst the profile of costing and cost control has been elevated since the global financial crisis – for example, through the establishment of independent budget offices or fiscal councils to validate expenditure and revenue estimates – there is very little detailed guidance on how cost information should be generated for routine utilization within an MTEF.

Recent World Bank-funded research on bottom-up costing practices adopted by four OECD countries helps to address this gap.[3] [4] Based on a detailed document review, as well as questionnaires and interviews with senior finance ministry officials, the study collates information on how costing practices currently operate within the MTEFs of Australia, Austria, Canada and the Netherlands. The countries represent a range of MTEF approaches. For example, Australia and Canada operate policy-based MTEFs that utilize rolling baseline mechanisms and elaborate program structures, whilst the MTEFs in Austria and the Netherlands work through a legal and policy-based architecture that utilizes ‘binding’ expenditure ceilings in the outyears, subject to tight parliamentary oversight and strong independent forecasting processes.

The study distinguishes two sets of costing practices. The first are guidelines, normally issued by finance ministries, to define cost concepts and methods of cost allocation. These guidelines use standardized costing definitions and methods. The second are centralized requirements, such as economic parameters and cost base assumptions, to ensure comparability in the costing of different programs. Whilst these two sets are usually intertwined, their separation for analytical purposes helps to illuminate the gap between the design and implementation of MTEFs.

The study made a number of observations on the relationships between program budgets, forward estimates practices and costing:

What are the key lessons that might be drawn from the research, especially in the context of costing for MTEFs in developing countries? First, even in countries with established MTEF budget processes, costing is an area that remains ‘under development’. Costing methods are surprisingly ad hoc, and used in a variety of ways. More to the point, some finance ministries consulted in the research conceded that the technical capacity required to cost programs within spending ministries is in urgent need of strengthening (as in Canada where a Costing Centre of Excellence has been established in the Office of the Comptroller General).

Second, care should be taken when building costing capacity within the MTEFs of developing countries. There remains an ever-present risk that the implementation of highly sophisticated and expensive cost methodologies – drawing on, say, activity-based costing or accruals data – can overwhelm human and systems capacity without necessarily delivering better information. Taking account of the variability and unevenness of costing practices in advanced economies, it may be more prudent, at least initially, for reformers in developing countries to focus on simplicity and greater consistency in the application of cost methods. Integrating basic – but comprehensible – cost information into MTEF budget routines might also help to reinforce among participants a broader awareness of the resource implications of their decisions. In this area, there is much to be said for the adage, attributed to John Maynard Keynes, that it is ‘better to be roughly right than precisely wrong’.

Finally, the role of bottom-up costing should be given more prominence in the practice-based literature on MTEFs.  Improved costing is seen as a building block for better informed decisions on multi-year resource allocation, and improved expenditure control, but the status of costing is often underestimated, and the capacity across government spending agencies to do it well frequently overestimated.

[1] Dr. Michael Di Francesco is a senior lecturer in public sector management and director of the case program at the Australia and New Zealand School of Government, Melbourne, Australia. Prior to this he led performance-based budget reform in the Australian State of New South Wales. He has also completed many PFM technical assistance missions with the IMF, World Bank and Asian Development Bank. He can be contacted at

[2] World Bank, 2013, Beyond the Annual Budget: Global Experiences with Medium Term Expenditure Frameworks, Washington DC, World Bank.

[3] M. Di Francesco and R. Barroso, 2015, ‘Bottom-up Costing within Medium Term Expenditure Frameworks: A Survey of Practices in Selected OECD Countries’, Public Budgeting & Finance, 35 (3), 44-67.

[4] M. Di Francesco and R. Barroso, 2015, Review of International Practices for Determining Medium-Term Resource Needs of Spending Agencies, World Bank Policy Research Working Paper WPS7289, World Bank, Washington DC.

Note: The posts on the IMF PFM Blog should not be reported as representing the views of the IMF. The views expressed are those of the authors and do not necessarily represent those of the IMF or IMF policy.