What is the appropriate role of line items in a performance budgeting system? How should the budget classification relate to the chart of accounts? What should be the role of the functional classification of expenditure? These are examples of questions that are made harder to answer – and which give rise to unnecessary confusion – because a crucial concept is missing from the conceptual toolkit of public financial management.
The missing concept is what can be called the expenditure control classification. This refers to the expenditure categories for which the government imposes quantitative limits, either in the budget law or by administrative order from the ministry of finance. If, for example, the government imposes a budgetary limit on the amount of capital expenditure of each ministry, then ministry capital expenditure is part of the expenditure control classification. If it also imposes a limit on how much is spent on each major capital project, then capital projects are part of the expenditure control classification. However, if the government leaves ministries free to shift money between projects within an overall capital expenditure allocation, then projects are not part of the expenditure control classification.
When the budget law appropriates amounts to ministries for specific purposes and those appropriations are limits that ministries are obliged to respect, the appropriations are part of the expenditure control classification. The same is true if the MoF sets such expenditure limits outside the budget law, via administrative instructions embodied in decrees, regulations or other instruments.
The terminology used to refer to the expenditure control classification varies between countries. In France, the term “limiting appropriations” (crédits limitatifs) is used, and these are mainly the broad categories of personnel and capital expenditure. In the United Kingdom, the terms “control totals” and “ring-fences” are used. The control totals essentially set limits on the capital expenditure and current expenditure1 of each ministry. The ring-fences refer to a small number of more detailed expenditure control categories, such as limits on the “administrative expenditure” of each ministry. Neither the control totals nor the ring-fences are set in the budget law. They are, rather, contained in “settlement letters” that the ministry of finance (HM Treasury) sends to each spending ministry.
Part of the Budget Classification
Thus defined, the expenditure control classification is a subset of the budget classification. Although it is widely considered to be a fundamental PFM concept, the budget classification is not clearly and consistently defined in the PFM literature. The most common meaning attributed to it is, however, that it covers multiple classifications of expenditure (and revenue) that are used for a wide range of purposes including planning, expenditure analysis and transparency.2 Used this way, the budget classification includes many categories of expenditure that the government does not use to control expenditure. It also includes detailed expenditure categories that individual ministries may control when they execute their own budgets, but that the government does not control. Defined in this manner, the budget classification includes elements such as the functional classification, the geographic classification, and the line-item classification of expenditure down to the most detailed level.
The expenditure control classification, by contrast, includes only those categories of expenditure that the government — i.e. the parliament and the ministry of finance — uses for control purposes.
Line Items
Using the concept of the expenditure control classification, the question of the role of line items under performance budgeting becomes easier to address, because it enables us to more clearly distinguish between the role of line items in financial reporting and their role in budget control. When performance budgeting proposes moving away from line-item controls, what it means is greatly reducing the use of line items in the expenditure control classification. For example, performance budgeting typically proposes abandoning all or most government-level controls over detailed line-item categories such as utilities and office supplies. The line-item controls that are retained are, typically, at a much more aggregated level – for example, controls over total goods and services expenditure, total personnel expenditure, or total capital expenditure, for each ministry. Within those broad categories, ministries are given the freedom to allocate their budgets as they wish. This has, however, nothing whatsoever to do with financial reporting. Expenditure is still accounted for and reported right down to the most detailed line-item categories. The line-item classification in the Chart of Accounts thus remains unchanged under performance budgeting.
In the most restrictive forms of traditional line-item budgeting, which still exist in certain developing countries, every detailed line-item category in the CoA is treated as a control category by the Ministry of Finance. The result that spending ministries are either unable to shift money between these categories, or can do so only after obtaining approval from the MoF. This is a source of multiple problems, ranging from poor performance to chronic under-execution of budget allocations. One factor contributing to this situation is an inability to distinguish clearly between line items as control categories and line items as accounting categories.
The Functional Classification
What about the role of the functional classification of expenditure – for example, in the form of the standard international COFOG classification? Here, the key point is that very few countries use the functional classification of expenditure as an element of the expenditure control classification. Functions are generally used only for informational purposes. The core role of the COFOG classification, for example, is to enable international comparisons of the functional composition of expenditure — a role that does not require it to be used for expenditure control purposes. The use of the functional classification for expenditure control purposes is confined to a minority of countries, including a number that inherited their budgeting systems from the former Soviet Union.
In summary, the functional classification is part of the budget classification — but not of the expenditure control classification — in most countries. This is a point that the PEFA diagnostic tool,3 and many PFM analysts, get wrong.
Relationship to the Chart of Accounts
This brings us to the question of the relationship between the budget classification and the Chart of Accounts (CoA). It is standard PFM doctrine that the budget classification should be fully “integrated” into the CoA, which means that every category of expenditure (and revenue) that is part of the budget classification should also be used by accountants to record transactions in the accounts. One of the consequences of this integration doctrine has been pressure on developing countries from international organizations to include the COFOG functional classification in their CoAs, so that each expenditure transaction is recorded by its functional classification when entered into the accounts. But this has been a mistake.
It is certainly not standard practice in advanced countries — in most of which government accountants do not record expenditure transactions by function in the accounts. Instead, the functional composition is estimated in an approximate way by a rough attribution of CoA categories — in particular, administrative categories — to functions (often with some percentage apportionment). This is considerably less demanding and serves the informational purpose. Including COFOG in the CoA adds an entirely unnecessary layer of complexity to government accounting — something that is particularly to be avoided in countries with very limited resources.
Although the proposition that the budget classification should be fully integrated into the CoA is incorrect, the expenditure control classification certainly does need to be so integrated. In other words, every category used to control expenditure should be either in the CoA or should be the sum of elements of the CoA. It is clearly not possible for the government to set a limit for any category of expenditure if it is not able to subsequently monitor actual spending on that category via the accounting system.
Entitlement Expenditure
Not every type of expenditure falls within the umbrella of the expenditure control classification. In particular, this classification does not cover entitlement expenditure under standing legislation — such as security expenditure — where the amount of expenditure is determined not by quantitative limits but by legal entitlements. Even if, as in countries such as Sweden and France, the budget law formally appropriates amounts for such expenditures, they are not part of the expenditure control classification because these “appropriations” do not limit expenditure if it turns out that expenditure required to meet entitlements under the relevant standing legislation is higher than the appropriated amount. In France, this is explicitly recognized by referring to such appropriations as “estimated appropriations” (crédits évaluatifs) — thus distinguishing them from the “limiting appropriations” mentioned in the box above. In the United Kingdom, expenditure aggregates that cannot be controlled by imposing quantitative limits because they include entitlement expenditure are referred to as “planning totals” to distinguish them from control totals.
To sum up, discussion of many of the questions that arise relating to the classification of expenditure would become clearer if the concept of the expenditure control classification became — like the Chart of Accounts and the budget classification — a part of the standard terminology of public financial management.
- More precisely, accrual expenses minus depreciation. ↩︎
- The IMF has defined the budget classification as “all coding schemes that are used to define both revenue and expenditure transactions at budget planning, budget approval, and budget execution” (Jacobs, Hélis and Bouley, 2009: 9). For similarly broad definitions see chapter 4 of OECD (2001), Managing Public Expenditure and Salvatore Schiano-Campo (2017), Government Budgeting and Expenditure Management. ↩︎
- PEFA indicator PI-4 requires, for an A rating, that “budget formulation, execution, and reporting are based on every level of administrative, economic, and functional classification using GFS/COFOG standards or a classification that can produce consistent documentation comparable with those standards.” ↩︎