Marc Robinson
Public Financial Management Results
Go to content

Is Program Budgeting Unworkable?

(Originally published 12 July 2011)

Program budgeting has many critics. One of the most prominent is Allan Schick, who has advanced the proposition that budgeting by programs is unworkable because it is inconsistent with budgeting by organizational unit. Schick’s argument (2007: 113-6) is based on three propositions, namely that:

  • An essential part of budgeting is the allocation of funds to organizational units.
  • Programs and organizational structures are “fundamentally antagonistic” – in other words, that allocating funds to programs is radically different from allocating it to organizational units.
  • It is impossible to allocate funds in the budget to both programs and organizational units.

From this, Schick concludes that program budgeting will only work if the “programs” are the same as organizational units, in which case the principle of programs based on product lines will in effect have been abandoned.

It can, however, be argued that only the first of these propositions is correct.

Under program budgeting, the government allocates funds in the annual budget to each ministry for that ministry’s programs, not for its internal organizational units. Of course, ministries must be able internally to allocate their budgets to their internal organizational units and must therefore be able to map program budgets to organizational units. This does not, however, mean that programs and organizational units must be identical. Consider the example of a ministry the internal organizational structure of which is based upon directorates, and sub-directorates, with both levels serving as budget holders. All that is necessary to reconcile program and organizational unit budgets under these circumstances is that the ministry explicitly identifies the allocation between sub-directorates of each program’s budget. In practical terms – although this is not absolutely essential  – this usually means that each sub-directorate will be identified with a single program. If this approach is applied, there is nothing to rule out several sub-directorates or even directorates being mapped to a single program (or sub-program). There is, indeed, nothing to prevent several sub-directorates from different directorates being mapped to a single program. In short, the need to explicitly link programs and organizational units does not mean that programs must be the same as organizational units.

The proposition that programs and organizational structures are “fundamentally antagonistic” is also hard to sustain. In fact, organizational structures in almost any government follow product lines to quite a large degree. Within an education ministry, for example, there will typically be separate directorates covering primary, secondary and tertiary education, and these will map naturally to primary, secondary and tertiary education programs. Similarly, within an environmental agency, there would certainly be separate directorates for nature conservation and pollution control, and these again would correspond directly to programs.

The same is at least as true of the allocation of responsibilities between different ministries or agencies, which in all countries largely follows product lines. In this context, Schick claims that the usual program budgeting practice of keeping programs within agency boundaries “robs program budgeting of its essential purpose”. He suggests, for example, that true program budgeting would require the police service, the courts, the prisons, the parole agency and a range of similar agencies be all grouped under the same program because they all aim to safeguard citizens against crime. But to suggest that program budgeting calls for the creation of giant programs of this type is to misunderstand its nature and objectives. The key objective of program budgeting is to facilitate improved expenditure prioritization at the government-wide level which means, concretely, decisions such as how much emphasis to place on prisons versus non-prison based correction, and how much emphasis to place on crime prevention versus catching criminals. A program classification will only facilitate such choices if it recognizes that, within the broad umbrella of anti-crime services, different agencies deliver a range of different product lines, and it is on the basis of these product lines that distinct programs should be defined. In other words, what program budgeting calls for is the creation of a number of specific programs such as “criminal investigation” and “criminal incarceration”, rather than one all-embracing anti-crime program.

Of course, organizational structure does not entirely follow product lines. The most obvious divergence between the two is that organizational structures always include units dedicated to the provision of internal support service units – such human resources management, IT and finances – which are by definition not ministry products. To the extent that organizational structure legitimately diverges from product lines, program budgeting usually accepts some compromise of the principle of product line-based programs in order to preserve a simple link between organizational structure and programs. In the case of support services, this is done by creating “support service” programs, in the full knowledge that these are not consistent with the pure program budgeting principle. The crucial point, however, is that such compromises of the principle of “product line” programs are strictly limited, and it remains the case that the great majority of programs are based on product lines.

Organizational structure sometimes diverges quite inappropriately from product lines and, where this is the case, program budgeting encourages – and should be seen as linked to – organizational restructuring. This is particularly the case under traditional, inward-looking civil service systems. Rational organizational restructuring consistent with a client-orientation should include, for example, the integration of separate organizational units which deliver closely related products. It should also in many cases include the elimination of organizational structures based on functions (i.e. professional competences/types of work process such as engineering) rather than products. Benoît Chevauchez (2007) makes the point that in France, precisely this type of organizational restructuring has been a key by-product of the new performance budgeting system legislated in 2001. This underlines the importance of viewing program budgeting not as an isolated reform, but as part of an overall “managing for results” reform package.

Claims that program budgeting cannot work because of organizational budgeting realities do not, then, stand up to closer analysis. The real challenge of making program budgeting work lies elsewhere – in ensuring that it is accompanied by the right type of performance information and that this information is actually used in the budget preparation process to influence the prioritization of expenditure.

Schick, Allen (2007), “Performance Budgeting and Accrual Budgeting: Decision Rules or Analytic Tools?”, OECD Journal on Budgeting, 7(2).
Chevauchez, Benoît (2007), “Public Management Reform in France”, in Marc Robinson, Performance Budgeting: Linking Funding and Results, Houndsmill: Palgrave Macmillan/IMF.
Back to content