« Stocktake of PFM Diagnostic Instruments | Main | Setting the Fiscal Deficit in India[1] »

October 30, 2017

Simplify Program Budgeting: Is There a Place for ‘Activities’ in a Program Classification?

Thinkstock_862564308_edited (002)

Posted By Khuram Farooq and Michael Schaeffer [i]

The traditional program budgeting structure is program, sub-program and activity. Can we simplify this structure? Is the ‘activity’ element of this classification really necessary?  Why ‘activity’?  The conventional argument is that ‘activity’ will be used to capture ‘projects’ or capital expenditures under sub-programs. This may be a valid explanation on paper. In practice, however, the concept of ‘activity’ has caused much confusion and difficulties during the implementation of program budgeting in several countries.

In Zambia, there were 3000 ‘activities’ under the Ministry of Health’s budget (FY 2011), with at least 5 subheads under each activity. This constituted around 15000 line-items! In Cambodia, there are 10,000 activities under the newly introduced program classification. This level of detail and complexity in the budget classification system is problematic on several counts, but the most fundamental issue is that it undermines the justification for moving to program budgeting.

One of the key justifications for moving to program or output based budgeting is that traditional budgets are too focused on details and inputs.  Budget documentation is often voluminous with (sometimes) thousands of pages of tables detailing inputs for each of the budget entities. Performance based budget systems, on the other hand are designed to shift the focus from ‘inputs’ to ‘outputs’ or ‘programs’ - the services to be delivered by Ministries, the targeted results for these services, and the reporting on the these results. In this framework, accountability for the use of government resources (taxpayer’s funds) is separated so that:

  • the government becomes accountable for allocating appropriate levels of funding to the ministries to meet the results; and
  • the line Ministries become accountable for the achievement of these results, and the effective management of the inputs entrusted to them by the government (through the budget).

This shift in accountability has been difficult in practice. The Ministry of Finance in various countries has been reluctant to give up their absolute control over detailed levels of inputs (often disguised as ‘activities’). The Line Ministries were tentative too, as often for the first time, they were being held accountable for service delivery performance.  

Too often, the response to these fears has been to continue providing detailed input and activity data in budget documentation and centralized expenditure control processes. The level of detail has often been so overwhelming that it diluted and to some extent undermined the focus on key service results, the very reason for the shift to performance budgeting systems.

Many other drawbacks have been experienced with respect to using activity level detail in the budget process:

  • It made the budget management process excessively cumbersome. During the budget formulation process, too much time and effort is spent on detailing every cost driver of an activity, at the expense of determining its effectiveness or whether the activity can contribute to service delivery or program outputs.
  • It made the budget execution inefficient. In Zambia, for example, it takes several weeks, every quarter, for the MOF to release funds for thousands of activities.  
  • Most of the information on ‘activities’ was either duplicative or unnecessary.
  • If activities are expenditure categories (e.g. repair of buildings, training, capital expenditures on buildings, works, equipment, etc.), these can be captured through economic classification;
  • If activities are internal work processes or the cost centers of the ministry (e.g. HR, Procurement, Budget, IT, Accounting, Internal Audit), as is often the case, these are unnecessary within the context of program budgeting. This level of detail is better suited for information within the ministry using spreadsheets.


But how to reflect capital and recurrent expenditures under programs?

One of the arguments supporting the use of ‘activity’ under program classification is that it reflects capital projects under programs. This objective is desirable. However, it can be achieved without using ‘activity’ - through mapping of spending units and projects to programs and sub-programs. Every ministry has spending units, financed either through recurrent or capital budget allocations (or a mixture of both). The thousands of schemes and sub-schemes, in which politicians are typically interested, can be treated as spending units. Since a program or sub-program should include information on recurrent and capital expenditures, all spending units, including schemes, should be mapped to the program classification in order to aggregate recurrent and capital expenditures.

What about spending units that contribute to more than one program?

Sometimes, it is argued that activities help in allocating costs to respective programs when a spending unit contributes to multiple programs. A spending unit at the top of the hierarchy (Minister’s office) or at the periphery (e.g District office) would present this scenario. An agriculture officer in a remote district will be contributing to multiple agriculture programs, including agriculture productivity and market access. This is one of the difficult scenarios under program budgeting. Such a scenario could be dealt in one of the two possible ways:

  • If an overwhelming part of the budget of this spending unit – for example, 80% of the budget - is being allocated from one program, this spending unit should be mapped only to that program.
  • If spending unit gets allocation from multiple programs in a mixed ratio – for example 30%, 20%, and 50% from three programs– there are two options to deal with this scenario:
    • For the purpose of budgeting and expenditure recording, create 3 separate codes for a spending units, and map each code in a one-on-one relationship with the respective program. The Maldives government has opted for this option under its budget reform. Or
    • Map such a spending units to general purpose ‘administration and management’ program

The underlying theme of the above options is that every spending unit code should be mapped to one program or sub-program. Using above design options under a programmatic budget approach is significantly simpler than using activity. Under this approach, the Ministry of Finance can approve the budget at program and/or sub-program level, while further distribution of budget to the spending units could be left to the ministry and program managers as shown below (See Figure 1). Budget execution controls and expenditure recording should be done at the spending unit level, and associated economic classification.

Figure 1: Simplified approach to program classification

Image 1

In sum, the introduction of activities in the budget classification system has created enormous difficulties in many countries. Many governments are grappling with these issues even after 15 years of experience with program budgeting. Eliminating ‘activities’ in the budget management system of these countries can simplify the process. Not only will the information duplicity be avoided, but the reforms will be refocused to the basic objectives of the program budgeting – achieving budget performance results.

[i] Michael Schaffer, Senior Public Sector Specialist, Governance Global Practice, World Bank and Khuram Farooq, Senior Financial Management Specialist, Governance Global Practice, World Bank.

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.


I can’t agree more with the assessment made in this excellent blog. It does seem beyond overconfidence to require capacity constrained governments to maintain a comprehensive activity breakdown of a program budget. As a traditional budget usually reflects an inefficient compromise of interests and power, such an interpretation of a program budget also provides bureaucrats with a few thousand extra arguments against performance informed reallocation and in favor of the status quo. Even OECD countries (perhaps with the exception of early LOLF implementation in France and South Korea) never realized ambitions to couple program budgeting with activity based costing for the entire government. Clearly there are parts of the public sector where recognizable public products can be distinguished relatively easy to serve as a stable basis for attributing activities and even do reliable cost accounting (e.g. forestry or inspection agencies often serve as example). For most of the public sector however the products distinguished in such systems tend to be artificial, meaningless to those in the primary process and subject to frequent change due to different political preferences. This is also the main reason why managerial use of accrual accounting information is reported to be disappointingly low. Is this reason to despair? I don’t believe so. Thinking about the alignment of activities and public sector goals and challenging cost levels with an activity and output based analysis are healthy for any government organization and can be useful for informing the allocation debate during budget preparation and budget execution phases. It is quite another thing to anchor down an incredibly detailed breakdown of resources based on a paper reality.

I disagree with this formulation and the elimination of the activity level. The activity level is where the costing of programs needs to occur. At a organizational unit level you will lose this. The alternative option does not seem correct to me. First the Health Ministry is the overarching organizational unit, but beneath that comes the spending budget unit in the organizational structure and it is here that the program/subprogram needs to be identified, such as the training or other activities that occur under this program/subprogram structure. The activities need to be broken down by the economic classification so that it is clear what is being spent and also to relate the inputs to the outputs that can only be disaggregrated at the activity level. Under the above formulation I believe you would lose this necessary level of program input/output relationship and also the ability of do program evaluation at at more granular level.

I don't agree with this formulation of eliminating the activity level in program budget. I think the examples and problems cited are really because of the lack of knowledge of how programs, subprograms and activities should be formulated in the program budgeting context. These examples tend to illustrate that program budgeting is being used on a line item budget basis, defining the minute level of spending as a basis for control of expenditures. In the formulation above, the spending budget unit falls under the program classification whereas it more appropriately belongs under the ministry level and then the program budget. The above formulation would seem to disguise or confuse the accountability for the programs to the implementation level. Also, activities can be formulated on a broader basis that just every action that is to be taken. Activities should be grouped into a broader category with the economic classification demonstrating the object of expenditure. Finally, activities are needed as the basis for costing of the inputs and outputs to a disaggregated level for program evaluation purposes. I don't see how that can be done if you eliminate the activity level. Too many broad activities would have to incorporated under the subprogram level and this would be difficult to identify specific inputs and their related outputs. This is the main purpose of program budgeting. I think the solution is better grouping and identification of the activity level rather than to do away with it.

I don't agree that you can eliminate activities in the program budgeting framework. The problem is not with activities but their use in the program budgeting framework. It is only at the activity level that costing can be applied. This is the real reason to have activities as well as the basis for identifying and measuring performance of the program and subprogram. You need this level of disaggregation in order to ensure that program budgeting is being effectively. I think the solution is to develop a better program/subprogram classification. What I see in my work in introducing program budgeting is that the programs/subprograms are really developed as activities rather than a classification of a program area. Clearer programs/subprograms would improve the development of the activities related to them.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.

Back to top of page
©2007 IMF. All Rights Reserved. About Us | Terms of Use
/************* DO NOT ALTER ANYTHING BELOW THIS LINE ! **************/ var s_code=s.t();if(s_code)document.write(s_code)//-->