It is often suggested that unit costs are the basic tool for performance budgeting. The proposition is that measuring the unit cost of public sector services (outputs) – or, according to some, activities – provides the best instrument for linking the funding provided to ministries to the results they are expected to achieve. In other words, unit costs are supposed to be used in budget preparation to calculate budget requirements as a function of the quantity of services to be delivered to the public. Monitoring budget execution then allegedly becomes as easy as seeing whether the planned quantities of outputs were actually delivered.
This exaggerated notion of the role of unit costs as the link between funding and performance is surprisingly widespread. I have in recent times visited two low income countries which are attempting to base entire performance budgeting systems on unit cost calculations. We’ve seen this before, and not only in developing countries. Australia and New Zealand made exactly this mistake, on a spectacular scale, in the nineties. 
Unit costs are indeed a powerful tool when selectively applied to the right types of public services. The most outstanding international example of a sectoral performance budgeting system based on unit costs in the “diagnostic related group” (DRG) hospital funding system. In education, where costs per student at particular levels of schooling tend to be relatively standard, unit costs can also be a powerful budgeting and performance management tool.
The catch is that unit costs can never be an across-the-board budgeting instrument, because there are many public services which do not have a stable unit cost. Take an extreme but illustrative case – police criminal investigations. The cost of one murder investigation can vary enormously from that of another, because the circumstances of the case differ. Another example: emergency services in a hospital, where the cost per treatment of patients tends to vary greatly and unpredictably, and which are for this reason excluded from DRG systems. It is easy to identify numerous other examples of such “heterogeneity”—that is, of costs varying because of differences in the effort required because of the circumstances of particular cases. But the problem doesn’t end there. How on earth does one fund an army, or a fire service, on the basis of unit costs of the outputs delivered? Such services are like insurance policies – government funds them not so much for services actually delivered, but to maintain capacity to provide crucial services if and when they are needed.
It is therefore quite inappropriate to seek to use unit costs as an across-the-board budgeting tool. Selective application is the way to go. Services which are suitable for the use of unit costing – such as hospital treatments and schooling – have in common is a considerable degree of standardization, which makes unit costs (relatively) stable. The selective use of unit costs as a performance budgeting tool for more standardized services makes good sense for governments seriously interested in improving public sector performance.
There is, however, a catch here. Using unit costs as a tool is a complex business. Good management accounting systems are required. Complex adjustments have to be carried out to allow for complicating factors such as regional cost differences. These and other technical challenges mean that one should be particularly selective and cautious in expanding the use of unit costs in developing countries with low capacity. It is a paradox that quite a few low income countries seek to introduce forms and tools of performance budgeting which are more complex than those used in the majority of OECD countries. Consultants undoubtedly have something to answer for here.
A technical footnote: Basing the budget on unit activity costs is not at all the same thing as basing it on unit output costs, for the simple reason that activities and outputs are not the same thing. “Activities” means things such as meetings, internal training seminars, and the production of ministry reports, whereas “outputs” means services delivered to the public (hospital treatments, education, etc). A budget based on activity costs therefore means, concretely, one in which the budget documents tell parliament and the public how many meetings, training sessions, internal reports etc they can expect to be delivered by each ministry with the funding it receives. But what parliaments and the public are interested in is not activities, but the outputs and outcomes which they generate.
1] M. Robinson “Output-Purchase Funding and Budgeting Systems in the Public Sector”, Public Budgeting and Finance, vol. 22 (4), 2002.
 See Part V of “Does Performance Budgeting Work? An analytic review of the empirical literature”, M. Robinson and J Brumby, International Monetary Fund Working Paper, 2005. Downloadable at www.pfmresults.com.
 M. Robinson, “Cost Information”, and “Purchaser-Provider Systems” in M. Robinson (ed.) Performance Budgeting: Linking Funding to Results (IMF/Palgrave Macmillan, August 2007).