2. Budget practices and governance in Peru

The Budget is a central policy document of the government, showing how annual and multiannual objectives will be prioritised and achieved. Budgets should present a full and accurate account of all government expenditures.

This chapter analyses the key elements of the overall budget process and budget governance in Peru in light of the OECD’s Good Budget Governance Practices (OECD, 2015[1]). The chapter provides an international benchmark against which to assess Peru compared to OECD countries, along with recommendations for aligning Peru’s practices with OECD standards.

Section 2.2 presents the regulatory and institutional framework for the Budget in Peru. Section 2.3 analyses the mechanisms for ensuring fiscal sustainability and the medium-term perspective, while Section 2.4 presents the budget formulation and approval process. Section 2.5 discusses a distinctive aspect of the Peruvian budget system, the significant and frequent budget modifications during the budget execution process. Another feature of the budget process in Peru is that it encompasses the national and subnational governments in the same budget process and Budget Law. Section 2.6, therefore, presents and analyses the specific challenges of subnational governments in the budget process in Peru. Section 2.7 analyses the link between budget expenditures and government priorities. The last section analyses accounting practices and control of public accounts.

The regulatory framework of the Budget is determined by the Legislative Decree (DL) of the National Budget System No. 14401 of September 2018, which amends the General Law of the National Budget System No. 28411 of December 2004. The objective of DL 1440 is to establish the principles, processes and procedures that regulate the National Budget System. The stated principles governing the Budget align with general OECD principles. For example:

  • Balanced budget: it is forbidden to include spending authorisations without corresponding funding.

  • Macro-fiscal balance: the Budget must respect the Fiscal Responsibility and Transparency Law and the Fiscal Decentralisation Law.

  • No earmarking: “the public funds of each entity are intended to finance all public expenditure provided for in the public sector budgets”.

  • Multiannual programming principle: “the budget process should be guided by the objectives of the National Strategic Development Plan and be based on previous years’ results and take into account the prospects for future years”.

  1. 1. Public Sector Budget Law

  2. 2. Law on the Financial Balance of the Public Sector Budget

  3. 3. Public Sector Debt Law.

These three laws are discussed and presented for the consolidated public sector (i.e. they include the national government, subnational governments and all public entities, such as universities, etc.).

The budget process in Peru is centralised in the Ministry of Economy and Finance (MEF), mainly under the stewardship of the General Directorate of Public Budget (DGPP) (Table ‎2.1). Public entities that receive a budget appropriation in the Annual Public Sector Budget Law are called “pliegos” in Peru; the term “budget units” will be used in this report.

The 1993 Political Constitution of Peru stipulates that the subnational governments have political and economic autonomy, approve their budgets, and have their own revenue. However, the Constitution also mentions that the public sector Budget voted by the National Congress contains a section for the decentralised bodies. This is a very unusual practice. In OECD countries, subnational government budgets are not included in the Budget Law voted by the national Congress. The Budget Law only presents the revenues constitutionally allocated to the national government as well as transfers to subnational governments. These can take the form of negative revenues in the form of tax sharing collected by the national government on behalf of subnational governments. Accounting then consolidates transactions internal to the public sector (transfers from the national government to the subnational governments) so that there is no double counting of the same expenditure. This allows for transparency in the revenues and expenditures constitutionally attributed to each level of government and the transfers between them.

In Peru, the consolidated public sector Budget is formulated and voted on without formulating and approving the budgets of the national and subnational governments independently. All revenue is considered as public sector revenue and no distinction is made between national government revenue, transfers and subnational governments’ own revenue.

In all countries, subnational governments must send their budgets to the Ministry of Finance for the consolidated public sector Budget. However, the budgets are first approved by the regional/municipal councils then consolidated into the public sector Budget, which guarantees subnational governments’ autonomy.

Integrating subnational government budgets into the Budget prepared by the MEF and approved by the national Congress allows the MEF to have very tight control over the budget execution at all levels of government, thus ensuring fiscal stability. It also ensures that the MEF has a complete picture of the revenues and expenditures of the public sector as a whole. This is not the case in all OECD countries. For example, in Greece, prior to the 2010 Kallikrates reform, there was no mechanism for ensuring the fiscal sustainability of the entire public sector. It took several months before the Ministry of Finance managed to consolidate execution data, so that deviations from fiscal targets were not identified in time to take corrective action (Box ‎2.2) (Moretti, 2019[3]).

However, including subnational budgets in the national Budget also creates challenges:

  • It makes the public sector Budget Law longer and more complex, as it must include a breakdown of each revenue and expenditure by level of government (aggregate), as well as for each regional and municipal government.

  • It is one of the elements that makes the subnational government budget process more complex, less transparent and less predictable (see Section 2.6).

  • It requires the use of institutional transfers, which represent a budget modification, instead of revenues transfers, which correspond to budget execution.

There are other mechanisms to ensure the financial sustainability of the public sector rather than merging national and subnational budgets into one document. Fiscal rules, for example, are a very important instrument for allocating deficit and debt authorisations between levels of government, since the public sector deficit and debt are calculated on the consolidation of the national government, subnational governments and social security (Vammalle and Bambalaite, 2021[4]). OECD countries also have mechanisms in place to monitor the execution of subnational budgets to ensure that they comply with the limits set in their budgets. Some countries have institutions to ensure that subnational budgets comply with the rules. For example, Greece recently implemented a major reform of the subnational government budgeting and reporting process, which both respects the principles of subnational government autonomy and reinforces the financial sustainability of the subnational sector (Box ‎2.2).

In all OECD countries, the national budget indicates the sources of revenue by type of revenue (e.g. value-added tax, fees and duties, personal income tax, corporate tax, etc.). This is usually done in the first section of the budget, which authorises the different types of taxes and fees and specifies their base and rate. However, all these revenues serve to finance all government expenditures, due to the principle of budgetary universality. Generally, countries have a budgetary category for funds received for specific purposes, such as financing by an international institution or a private actor of an investment project or a specific expenditure (such as the “fonds de concours”, or competitive funds, in France). However, these are often small compared to total revenues.

In Peru, the budget not only gives estimates of revenues from different funding sources, it also allocates expenditures according to these funding sources (Table ‎2.2 and Figure ‎2.1). Annex 4 of the budget shows the expenditure distribution by level of government, budget unit and source of financing.

This creates a high level of rigidity, as it requires a budget amendment to transfer expenditures from one source to another if there is, for example, an error in the revenue estimation per source. It creates an additional administrative burden on entities to link their expenditures to their revenue sources and creates complications in treasury and account management, as each expenditure must be tagged to a revenue source rather than simply coming out of a Single Treasury Account that would pool all revenue sources (see Chapter 3).

In addition to these administrative complications, the way differences between the amount of revenue estimated in the budget and the revenue actually collected are treated varies depending on funding source. This differential treatment creates incentives for different entities to under- or over-estimate different revenue sources (see the discussion in Section 2.6.1). Indeed, up to and including the 2022 Budget, any directly collected revenue that exceeded the amount estimated in the Budget Law it could be spent at the discretion of the head of the entity without the need to go through Congress again. A possible consequence is that “modified” budgets do not always correspond to the distribution of spending and priorities established in the Council of Ministers and discussed and approved by Congress. This is not in line with the OECD Recommendation of the Council on Good Budgetary Practices, which states “closely align budgets with the medium-term strategic priorities of government” (OECD, 2015[6]).

From the 2022 Budget, the national government’s directly collected revenue collected above budget will be returned to the Treasury. This is a first step towards reducing the ratio of funding sources to spending in practice, but it does not lead to administrative and treasury management simplification.

It is worth mentioning that the funding sources used in Peru do not correspond to the generally used categories of public revenues (Box ‎2.3). In addition, the Peruvian budget considers resources from official credit operations as a source of financing of the same nature as other revenues. International public accounting practices, and in particular the System of National Accounts, calculate the balance in the non-financial accounts (net lending/net borrowing) as the difference between total revenues and total expenditures then describe in the financial accounts how this deficit will be financed (or surpluses will be reversed). According to the System of National Accounts, resources from official credit operations are not normally included in the income section).

For each budget unit, the budget is presented by generic expenditure, expenditure category, level of government, function and source of funding, and different combinations of these categories (Table ‎2.3). This implies a very long and detailed budget bill, but one that is essentially focused on inputs rather than results.

The bill submitted to Congress comprises the Public Sector Budget Law (approximately 80 pages and 5 300 lines), which includes 13 annexes totalling more than 2 800 pages and approximately 84 000 lines).2 This reduces the clarity of the Budget Law and creates rigidities when it is voted on at a very high level of detail.

Accepted good practice in OECD countries points towards short budget laws, with relatively high-level allocations and an emphasis on outputs rather than inputs. These laws are supplemented by annexes, which provide more detailed information on how expenditures, targets and indicators are allocated, but they are usually not part of the Budget Law and can therefore be amended without amending the Budget Law. A 2012 study, for example, showed that ten countries have budgets with less than 300 lines (OECD, 2014[8]).

To some extent, this meant that the public budget approved by Congress did not constitute expenditure ceilings for the entities but allocated ordinary revenue for each expenditure, with total expenditure depending on the entity’s level of directly collected revenue. This was clearly stipulated in the previous General Law on the National Budget System (No. 28411): “the limits of the budget appropriations are constituted by the estimated revenues that the entities expect to receive, as well as the public funds that have been determined and communicated to them by the Ministry of Economy and Finance”.

This practice allowed limiting the fiscal risk for the national government, since the amount of allocated regular revenue is determined. However, it also meant that budget discussions were not held on the amounts of each entity’s expenditures, since these depended on the level of each entity’s own revenue collection and on the estimated levels of these. This reduced the usefulness of the budget as an instrument for prioritising public spending and politicised the estimation of directly collected revenue, when revenue estimation should be a purely technical exercise.

It should be noted that from 2023 onwards, directly collected revenue from national government entities will be considered as ordinary revenue, so this challenge should be resolved for national government budget units.

According to OECD good practice, the national budget sets spending limits for each programme/allocation, regardless of each entity’s own level of revenue collection. This is generally achieved automatically by not distinguishing between funding sources. However, there are certain cases where public entities are allowed to collect and retain fees and payments (in particular, where the costs to the entity depend on the level of service, such as the issuance of passports). However, there are usually rules to avoid creating incentives for entities to over-collect these fees and taxes or to under-declare their estimates (Box ‎2.4).

The situation is different for subnational governments. As mentioned above, in OECD countries, subnational government budgets are separate from the national government’s budget. The common practice in OECD countries is that transfers from the national government to subnational governments do not depend on the level of the subnational governments’ own revenue collection.3 Therefore, subnational governments’ level of spending does depend on the level of own revenue collection.

Based on the analysis presented in the previous sections, this report proposes the following recommendations (Table ‎2.4).

In the late 1990s, after several decades of macroeconomic instability caused by a lack of prudent fiscal and monetary policy management, Peru began to develop an institutional framework to ensure fiscal stability (Box ‎2.5). Today, the regulatory framework for fiscal responsibility is based on two instruments:

  1. 1. Legislative Decree 1276 of 2016 establishes four fiscal rules for the non-financial public sector:4

    • Debt rule: Total gross debt of the non-financial public sector ≤ 30% of gross domestic product (GDP).

      Exceptionally, in cases of financial volatility and provided that the other macro-fiscal rules are complied with, public debt may have a temporary deviation of no more than 4 percentage points of GDP.

    • Economic performance rule: Non-financial public sector fiscal deficit ≤ 1% of GDP.

    • General Government Non-Financial Expenditure Rule: The real annual growth rate cannot deviate by more than 1 percentage point from the 20-year average of real annual GDP growth.

      The average is calculated using the real GDP growth rates of the 15 years before the elaboration of the multiannual macroeconomic framework, the estimate of the fiscal year in which the multiannual macroeconomic framework is prepared, and the projections for the four years thereafter.

      The real rate is calculated using the Consumer Price Index of Metropolitan Lima according to the multiannual macroeconomic framework.

    • General Government Current Expenditure Rule: General government current expenditure (excluding maintenance expenditure) must respect the above rule (i.e. offset current expenditure growth cannot be offset by reducing capital expenditure growth).

  2. 2. Legislative Decree 1275 of 2016 adds two fiscal rules for subnational governments:

    • Regional or Local Government Total Debt Balance Fiscal Rule: Total debt balance ≤ annual average of total current revenues for the last four years.

      Where the total debt balance comprises the balance of financial liabilities, debt owed to government entities and real debt to private pension fund managers, and total current revenue comprises tax, non-tax revenue and current transfer revenue.

    • Current Account Savings Fiscal Rule: Total current revenue ≥ total current non-financial current expenditure.

As in most countries, macro-fiscal rules can be modified exceptionally in the case of disaster or significant extreme shocks. DL 1276 clearly states the procedure to be adopted and that this temporary modification should explicitly state the gradual return to the rules. This clause has recently been used as a consequence of the national emergency caused by the COVID-19 pandemic for fiscal year 2021.

The MEF is in charge of calculating fiscal rules and monitoring their application.

In the case of the rules for the non-financial public sector, slippages must be compensated for the following fiscal year. In the case of regional and local governments, the MEF can adopt corrective measures in case of non-compliance. These include, for example: preventive monitoring, prevention from entering into new short-term debt operations or from entering into a new public-private partnership contract. Subnational governments subject to corrective measures must submit a “commitment to comply with fiscal rules” to the MEF and the Comptroller General of the Republic. The commitment must be approved by the regional or municipal council and explains the measures taken to avoid further non-compliance.

The MEF can also provide technical assistance to implement good practices in fiscal management to subnational governments that request it.

The MEF publishes regular monitoring reports on subnational governments’ public finances and compliance with the rules (four quarterly reports and one annual report each year).

In many OECD countries, the Ministry of Finance makes a fiscal space calculation resulting from the application of fiscal rules. The revenue projections and the various rules generate a firm expenditure ceiling for the public sector and for each level of government. In countries with a top-down budgeting process, the budget process starts with the fiscal space calculation and the political decision on the level of deficit and expenditure (which cannot be higher than stipulated in the fiscal rule but can be lower). This is usually clearly shown in the first pages of the draft and then of the public Budget Law.

In Peru, the opening institutional budget (and the modified institutional budgets) do not clearly show the level of the deficit because the resources from ordinary lending operations are presented together with the funding sources. However, there is a difference in nature between tax revenues and debt financing. In most countries, credit operations are presented as financing the net deficit (the difference between total revenues and total expenditures). In the System of National Accounts, net borrowing/net lending is equal to the difference between total revenues and total expenditures.

The MEF produces three multiannual documents for the budget process. The medium-term perspective is thus taken into account and efforts are made to make economic projections and expenditure estimates. However, the revenue and expenditure assumptions and estimates used in the annual budget process are not consistent with these documents (Table ‎2.5).

The OECD has identified good practices for implementing a medium-term expenditure framework (Box ‎2.6) and analysed the practices in those countries that best implement them (Table ‎2.6). In several OECD countries, the medium-term budgetary and/or fiscal framework is discussed and approved by parliament. It is in these discussions that the parliament analyses and decides on the allocations of available funds among different sectors and priorities. It should be noted that it takes a long time to reach these levels, and it has taken decades for these countries to develop and refine these practices.

Recommendations for strengthening fiscal sustainability and the medium-term perspective in the budget are presented in (Table ‎2.7).

The budget calendar in Peru is not very clearly institutionalised, and most of the major milestones have dates that are aspirational, but not legally binding (Table ‎2.8). Having and respecting a fixed calendar is a crucial element of a budget system. On the one hand, it allows the different actors to plan their workload during the year. On the other hand, it also allows the discussion process to be limited in time, as all actors know at which point in time they can discuss which issue. This also increases the transparency and fairness of the system by giving all actors the same opportunities and time to discuss.

OECD countries usually have an institutionalised budget calendar with clear and legally binding dates for the delivery of key documents (see the case of Greece in Box ‎2.7). In European Union countries, these guidelines are given at the European level, and all countries must follow them (Table ‎2.10).

The budget process in Peru is initiated in January or February by the DGPP, which issues the Multiannual Expenditure Programming Directive. This document indicates to the budget units the calendar for the year and what they must do during the year. It is a technical document with a lot of administrative details (forms to be filled in, information to be entered into the system, etc.). The DGPP tries to issue this document in December, but it is often not issued until January or February.

In OECD countries, it is common for the budget process to start by issuing an “annual circular” or annual directive. Generally, the General Budget Law indicates when this directive should be issued. These directives usually contain information on macroeconomic assumptions, fiscal targets and policy priorities. In many countries, it also shows the expenditure ceilings for each ministry, as set out in the medium-term fiscal strategy. It also indicates deadlines, formatting guidelines and documents to be sent to the Budget Directorate by the various budget units (see, for example, the case of Greece in Box ‎2.7).

In Peru, there is no mandatory date for the publication of the Multiannual Expenditure Programming Directive. The directive does not include information on the macroeconomic situation or fiscal targets, nor guidelines on the target for expenditure increases, fiscal effort, etc. Budget units do not, therefore, have this information to start calculating their baselines, prioritising their actions or planning their budgets for the following year.

The first stage of the budget process in Peru consists of estimating the different revenues to be collected. These estimates are made by the General Directorate of Macroeconomic Policy and Fiscal Decentralisation based on information provided by the National Superintendency of Customs and Tax Administration (estimated revenue collection), the General Directorate of Public Treasury (estimated financial market and credit operations), and the Central Reserve Bank (estimated growth, interest rate and inflation rate). Directly collected revenue estimates are made through discussions between the DGPP and the different budget units.

The allocation of expenditures according to funding sources implies a very complicated process since it is not sufficient to estimate the overall evolution of revenues in each type of source, but it is necessary to have a very precise estimate of each entity’s directly collected revenue. Indeed, the level of expenditure allocated to each entity will depend on the directly collected revenue estimate used in the public budget. If the budget did not link expenditures to funding sources, it would be sufficient to have an overall estimate of directly collected revenue, and expenditure allocations would not depend on these, but rather on the evolution of baselines and political decisions to prioritise new expenditures. By linking spending authorisations to the level of directly collected revenue, estimating directly collected revenue also becomes a political process.

Revenue estimation is an essential element of the budget process; the level of authorised expenditure depends on it. In many countries, there is a tendency to overestimate these resources to increase the level of expenditure, which results in higher than desired deficits. To avoid conflicts of interest and improve the quality and reliability of revenue estimates, many OECD countries have designed processes to control the quality of revenue estimates (Box ‎2.8).

The inputs used by the General Directorate of Macroeconomic Policy and Fiscal Decentralisation to make its ordinary revenue estimates are not published, and there is no explanation of how the estimates are calculated based on the various inputs nor subsequent analysis to explain the differences between the estimates and the levels actually collected. Directly collected revenue estimates are negotiated with budget units and there are incentives to underestimate them (see next section).

The budget units must make their expenditure requests based on the multiannual budget allocations. However, these multiannual budget allocations do not reflect credible expenditure ceilings but are de facto considered as the starting point for discussions that will lead to increased expenditure allocations. The budget units prepare their requests based on their estimated needs without having a credible spending ceiling. In June and July, the budget units and subnational governments conduct political discussions to obtain more resources to finance their new priority programmes. Depending on the evolution of the revenue estimates, the MEF (DGPP) accepts additions to the amounts established in the multiannual budget allocations.

This type of budget process, where revenues and expenditures are discussed simultaneously, corresponds to the traditional practice of bottom-up budgeting in OECD countries. Today, most OECD countries have moved towards a system of “top-down budgeting”, which allows for better control of total spending, improves the quality of budget proposals, and reduces the wear and tear associated with lengthy and extensive political discussions to formulate a sustainable budget (Box ‎2.9).

Congress has three months to review, discuss and amend the draft Budget Law. It is allowed to comment on proposed and prioritised expenditures but cannot increase the total amount of expenditures. This is in line with OECD good practice.

Sectoral ministers and subnational governments substantiate their spending needs before the Congressional Budget Committee. Annex 5, which details the investment projects that will receive budget funds, usually receives the most attention and modifications. However, as discussed in Chapter 5, there are no limits or legal framework governing the changes that congresspeople can make to the list of projects included in Annex 5. In particular, while the government’s Annex 5 proposal is developed based on a rigorous prioritisation and quality control process, the projects added by congresspeople during this phase of the budget discussion do not have the same technical and quality controls. In principle, the only limit on members of Congress’ ability to amend Annex 5 is not to increase total spending. In practice, however, it is common for total spending to be increased during this phase of the budget process.

It is very unusual for subnational governments to have to submit their budgets and substantiate their spending needs to Congress. There are always discussions to increase transfers from the national government to the various subnational governments, but not the substantiation of the subnational governments’ budgets.

The budget approved by Congress is called the “opening institutional budget”, and it is established that it represents a floor for expenditures and will be modified throughout the year. In most cases, Congress does not review or authorise these modifications. Some of these modifications have been pre-approved in the opening Budget Law, but as the voted budget is consolidated between levels of government, what in other countries would be recorded as expenditure execution (transfer from the national government to a subnational government), is recorded as an “institutional transfer of item” in Peru, which corresponds to a budget modification.

In OECD countries, Congress must authorise all amendments to the Budget Law. However, in many countries, only the overall allocations per ministry/budget unit or programme are binding in the Budget Law. More detailed information on how each ministry/budget unit/programme’s expenditures are broken down is given for information purposes only, and ministries/budget units/programmes can reallocate expenditures from one line to another within their budget allocation without requiring congressional authorisation. After the fiscal year, each entity issues accounts and reports on how it spent the money along with the results achieved.

In most OECD countries, the draft budget bill and the Budget Law indicate the values of revenues and each allocation for the previous year (estimated) and for year t-2 (final values). This makes it possible to analyse the evolution of different expenditures during discussions and to verify whether this evolution is consistent with government priorities. Many countries also indicate the amounts foreseen in the medium-term budget framework for years t+1, t+2 and sometimes even t+3.

Based on the analysis presented above, this report proposes the following recommendations (Table ‎2.12).

An originality of the budget system in Peru is that the budget discussed and approved by Congress is called the “opening institutional budget” and is considered as a spending floor. All actors (the MEF, sectoral ministries, public entities, subnational governments, etc.) know and expect this opening budget to be modified during the year.

All countries must make adjustments and corrections to their budgets during the budget year (Box ‎2.10). The most frequent causes of budget modifications in OECD countries are: changes in economic forecasts (modifying revenue estimates, for example), increases in mandatory expenditures, implementation of fiscal stimulus measures, response to emergencies (e.g. natural disasters, COVID-19), funding of new policy initiatives or reallocations of funds without an increase in total spending (Figure ‎2.5).

In Peru, modifications are larger and more frequent than in OECD countries. For the national government, 45 modifications to ordinary revenue made by supreme decrees were counted in 2018, for example. Adding up all sources of funding, the absolute value of the balance of modifications5 represents a net increase of 10% of the initial expenditure level for the national government (Table ‎2.13 and Figure ‎2.6). In addition to increasing total spending, budget modifications also change the composition of the budget and can change the weight of each sector in total public spending. A recent study by the Office of the Comptroller General shows that modifications made during budget execution represent a deviation of 54%, on average, from the amount approved for the period under evaluation (Shack and Rivera, 2022[16]).

An analysis of regional and local government budget modifications shows the following:

  • In 2018, budget modifications accounted for almost half the initial budget for regional governments and more than the initial budget (136%) of local governments (see Table ‎2.13).

  • For regional governments, about one-third of the modifications correspond to resources from official credit operations and another third to increases in ordinary revenue. Almost all of the increases in the ordinary revenue sources correspond to revenue transfers from the national government to the budget units. The most important transfers come from the Ministry of Education, the Ministry of Health, the Ministry of Transport and Communication, and the MEF.

  • The dates of the supreme decrees authorising budget modifications show that as early as January-February there are modifications. Most of the modifications are then concentrated in July-August, but modifications are still being approved until December (Figure ‎2.7).

When comparing budget execution levels with the level of the last modified institutional budget, budget execution levels appear rather low, especially in regional and local governments (Figure ‎2.8). However, all levels of government execute budgets above their opening institutional budget. Indeed, subnational governments continue to receive additional resources in the last months of the year (Figure ‎2.8), which implies that it is almost impossible to execute these expenditures before 31 December. This is even more relevant for investment spending since executing an investment project takes time (see the discussion in Chapter 5 on infrastructure).

The extensive and frequent use of budget modifications breaks several fundamental principles of good public budgeting practice (OECD, 2015[1]):

  • Principle 1: Manage budgets within clear, credible and predictable limits for fiscal policy.

  • Principle 2: Align budgets closely with the government’s medium-term strategic priorities.

  • Principle 5: Facilitate an inclusive, participatory and realistic debate on budget alternatives.

Indeed, the real spending limits are not known when discussing trade-offs between different sectors and spending areas, as the budget discussed and voted does not represent what will be spent; actual spending may not be closely aligned with the government’s strategic priorities; and the political and parliamentary debate takes place on amounts that do not represent what will be spent. This can lead to revenue being spent on initiatives other than those prioritised in the initial budget.6

Budgetary changes during the year modify not only the volume of total expenditure but also its composition. Indeed, annual budget planning is based on the opening institutional budget. But the opening institutional budget may be too small for certain priority expenditures, so that revenue is allocated to other, lower priority expenditures. Successive increases in spending may all be too small to finance that priority, so that all revenue is allocated to lower priority spending. However, it is possible that, at the end of the year, looking at the amended budget as a whole, it would have been sufficient to fund the initial priority. But by not having the full information from the formulation of the budget, it is impossible to finance some priority expenditures, despite having the resources.

Budget modifications must be reviewed and authorised by the MEF, and an entire team (DG Thematic Budget) is mobilised to carry out this work. This comes at a high cost and absorbs human resources that could be working on more strategic issues related, for example, to the quality of spending, its efficiency, etc.

Finally, in many cases, the modalities of discussion and validation of budget modifications do not require further discussion and approval by Congress or the MEF. In particular, when it comes to directly collected revenue collected in excess of the budget, the allocation of spending is at the discretion of the head of the entity, which is not in line with principles of spending transparency.

To close the gap between the opening institutional budget and the modified institutional budget, it is necessary to understand how budget modifications are generated.

There are several sources of budget modifications: large and systematic errors in the estimation of revenue, leading to actual revenue collection in excess of the estimates used in the voted budget; late incorporation of carry-overs; late incorporation of canon and royalties; and line item transfers. This section describes each of these sources of changes and proposes an explanation of why this occurs.

There is a systematic bias in underestimating revenues in the opening institutional budget in Peru. Between 2016 and 2020, total revenues collected (the sum of all funding sources) were, on average, 20% higher than the revenues estimated in the opening institutional budget due to a significant underestimation of directly collected revenue (Figure ‎2.9). The difference between the estimate and the execution is smaller for ordinary revenue, and there is no systematic bias.

There are several explanations for this systematic bias:

  • Part of the public revenues in Peru depends on exogenous elements, which are difficult to estimate, such as international prices of natural resources, terms of trade, etc.

  • A very prudent estimation of revenue ensures the fiscal sustainability of the public sector. By not allocating resources before revenues actually enter the Treasury, the MEF ensures that the kind of crisis that Peru experienced in the 1980s is not repeated.

  • Budget units have incentives to underestimate their directly collected revenue. Indeed, up to and including the 2022 Budget, directly collected revenue collected above the levels voted in the budget was freely available to the incumbent. This creates the risk of a misuse of resources, and allows entities to retain a higher level of resources, even if it is not necessarily aligned with the government’s priorities.

  • There is no mechanism to control the assumptions used in the budget preparation. The MEF (DGPP) receives information from the National Superintendency of Customs and Tax Administration, the Central Reserve Bank, budget units, subnational governments, etc., but no alternative estimates to those made by the MEF are published. Peru has an independent fiscal institution, the Economic and Fiscal Council, but it is not within its competence to issue an opinion on the assumptions proposed by the MEF for the budget.

Traditionally, the practice in Peru is to wait for definitive information on the amount of certain revenues before integrating them into the budget rather than using an estimate of them (as is done for other revenue).

Another source of resources whose amount is only known during the year are rewards for meeting targets: “budgetary programmes” have targets, and when a budget unit participating in one of these programmes meets its target, it is entitled to receive a reward. The practice in Peru today is that this award is paid the same year the target is met, therefore resulting in a budget modification.

The calculations made by the MEF to determine each municipality’s FONCOMUN allocation are also made after the beginning of the year. Therefore, local governments receive information on the amount of FONCOMUN allocated to them during the current year and must make a budget modification to integrate them. As the subnational government budgets are consolidated in the public sector budget, this implies a budget modification of the public sector budget.

As mentioned in previous sections, in OECD countries, subnational government budgets are not consolidated in the national government budget. Consolidation is an accounting exercise performed after the approval of the various budgets. When a national government budget unit wants to execute its expenditure by making a transfer to a subnational government, the transaction is booked as expenditure for the national government budget unit and as transfer revenue by the receiving subnational government. This transaction is then consolidated in the general government (or public sector) accounts, which consolidates all public revenues and expenditures cleaned up by transfers between levels of government so as not to artificially generate a higher level of expenditure than is economically appropriate. A consolidated budget can also be produced in the same way, cleaning up such transactions.

In Peru, the approved budget consolidates levels of government and public entities; unconsolidated national government budgets are not presented. Therefore, Peru uses a very unusual mechanism: "line item transfers”. This consists of a budget modification that reduces the budget of one budget unit and increases the budget of another. This makes it possible to have a consolidated budget at all times instead of having to wait until the end of the year to consolidate the accounts to obtain consolidated information. This practice mixes two fundamental and distinct concepts: budget and budget execution.

These transfers of items occur for various reasons in Peru, including:

  • Vote of a consolidated budget, which does not allow transfers between public entities to be recorded as budget execution.

  • Lack of planning by line ministries: Line ministries do not plan their expenditures sufficiently in advance to include all of their transfers for the year in the opening institutional budget. This is partly due to a lack of planning capacity, but also because of the vicious circle of not having clear expenditure ceilings, which would define the limits for such planning.

  • Used as an instrument to control the execution of expenditure: line item transfers are used by line ministries as an instrument to control the execution of expenditure by the receiving budget unit (in particular subnational governments): they are granted when the line ministry deems that certain necessary conditions are met.

The transfer of items in Peru generates two complications:

  • As mentioned above, line item transfers constitute a confusion between the budget and execution, requiring the budget to be modified when in fact, it is simply being executed (in particular for modifications pre-authorised in the Budget Law).

  • These transfers occur late in the year, which does not allow the receiving budget units to execute them correctly, thus creating carry-overs (see below).

Incorporating carry-overs is particularly important for local and regional governments. For example, in 2018, directly collected revenue increased by 23% compared to the initial budget (PEN 180 million [Peruvian soles]), of which more than 60% (PEN 109 million) corresponds to rolling forward the carry-overs (MEF, 2019[17]).

As shown in Figure ‎2.9, budget execution is always higher than the opening institutional budget, but more often than not, it is lower than the modified institutional budget, which generates a carry-over. The main sources of budget increases during the year that contribute to generating balancing balances are the incorporation of additional resources throughout the year, incentives to generate balancing balances (because these are spent the following year "at the discretion of the Head of the Entity"), and the incorporation of balancing balances themselves.

The practice in Peru today is to wait for the final accounts with the exact information of the previous year's balance balances before modifying the budget to incorporate these balances. Indeed, it is very difficult to estimate balance sheet balances, given for example that revenues continue to increase until the last month of the year - and therefore, there is no information on the resources that are potentially not going to be executed. From 2019, a rule obliges pliegos to incorporate balance sheet balances in the first quarter of the year. The resources that are not incorporated by that deadline are transferred to the Treasury to finance the budget of the current fiscal year.

OECD countries tend to heavily regulate the carry-over of balance sheet balances. All OECD countries have rules that allow to some extent a "carry over" of unused funds to the following years. In general, there are limits on the possibilities for carry overs (Box ‎2.11).

There are also budget modification mechanisms to reduce the level of expenditure in cases where the amount of revenues collected is lower than the projections used in the budget. So far, they have hardly ever had to be used. However, it is an important mechanism that could play a greater role if the bias to use revenue underestimation were reduced.

In recent years, two measures have been adopted to close the gap between the opening institutional budget and the modified institutional budget, and to modify the incentives for the actors involved in the budgetary process to increase transparency and efficiency. These are:

  • From 2019, budget units are authorised to incorporate carry-overs up to the first quarter (DL 1441).

  • From 2023 onwards, national government budget units’ directly collected revenue are constituted in budgetary terms as ordinary revenue (DL 1441).

As detailed in the previous section, Peru makes significant use of budget modifications as an instrument to control budget execution and ensure that resources are sufficient to cover expenditures.

The annual commitment schedule is a short-term public expenditure programming instrument for all funding sources, which makes it possible to reconcile the cash programming of revenues and expenditures with the real financing capacity for the respective fiscal year, within the framework of the fiscal rules in force.7 The cash programming of revenues is provided by the General Directorate of Public Treasury, in compliance with fiscal rules and the multiannual macroeconomic framework.

This instrument is determined on 31 December of the year prior to the fiscal year and reviewed and updated on a quarterly basis by the DGPP based on information provided by the budget units.

The annual commitment schedule is a very simple document, which indicates the authorised expenditure ceilings for each budget unit of the national government, as well as for each of the regional and local governments. The annual commitment schedule indicates global amounts per budget unit without distinguishing between funding source or expenditure type.

The initial annual commitment schedule is usually lower than the opening institutional budget amount, and amounts increase each quarter as the budget changes, always remaining below the modified institutional budget. Executed expenditure is systematically lower than the annual commitment schedule (Figure ‎2.11).

The methodology for calculating expenditure commitments in the annual commitment schedule lacks transparency, and the MEF does not provide an explanation of how the amounts in the annual commitment schedule relate to the amounts in the various documents (opening institutional budget, modified institutional budget, multiannual macroeconomic framework, etc.).

Currently, it is mainly used as an instrument to monitor spending commitments. However, this could be a very interesting instrument to control expenditure execution and ensure that commitments do not exceed the revenue collected.

The modern trend in OECD countries is for budgets voted by Congress to be relatively short, setting strict expenditure ceilings (to ensure fiscal sustainability), allocated at a high level (by sector/mission/ministry) and indicating the objectives to be achieved with those expenditures. The detailed breakdown of funds voted by Congress between different programmes and budget lines is often done by the executive (Ministry of Finance, Presidential Office, etc.). OECD countries use various types of instruments to monitor budget execution and ensure that targets are achieved and expenditure ceilings met (Box ‎2.12).

Another important trend in OECD countries is to separate the payment authorisation functions (carried out by sectoral ministers) from the accounting and payment functions (carried out by a director of financial services). This reform is sometimes described as the creation of “a small finance ministry in each sectoral ministry”. In France, for example, each sectoral ministry has a public accountant responsible for the Service du Contrôle budgétaire et comptable ministeriel. In Greece, the creation of directorates general of financial services in line ministries is unanimously considered one of the major successes of recent reforms, having achieved its objectives: improving budgetary capacities, streamlining the payment process between ministries and improving communication with the Ministry of Finance. This reform is also considered essential for implementing another reform: top-down budgeting (Box ‎2.13).

Table ‎2.14 presents recommendations based on the analysis presented in the previous sections.

The 1993 Political Constitution of Peru states that regional governments, as well as provincial and district municipalities, have political, economic and administrative autonomy in matters within their competence (Articles 191 and 194). Their representatives (president of the region, vice-president and members of the regional council, mayors and councillors of local governments) are elected by direct suffrage (Articles 191 and 194) and are thus accountable to the citizens. The Constitution also establishes that regional and local governments approve their budget, formulate their development plans, and manage their assets and revenues (Articles 192 and 195). Finally, the Constitution indicates the subnational assets and revenues, which include specific transfers allocated to them by the annual Budget Law, certain taxes, certain resources allocated from the Regional/Municipal Compensation Fund, royalty revenues and resources from official credit operations (Articles 74, 193 and 196).

According to the Peruvian Constitution, the economic and financial administration of the state is governed by the budget approved annually by Congress. The public sector budget contains two sections: the central government and decentralised bodies (a term that designates, among others, subnational governments) (Article 77).

As mentioned in the previous section, the budget discussed and voted on in the national Congress is consolidated between the national and subnational governments. No non-consolidated budgets are presented that clearly show the revenues of each level of government or the transfers between them. In particular, Peru’s Budget Law considers all revenues as public sector revenues, without distinction between levels of government. Article 77 of the Constitution mentions that “the budget allocates public resources equitably, its programming and execution respond to the criteria of efficiency of basic social needs and decentralisation. The respective districts shall, in accordance with the law, receive a fair share of the total income and revenues obtained by the state from the exploitation of natural resources in each area as a canon”. But it does not indicate how these concepts are measured or secured. Annex 4 of the annual Budget Law presents a breakdown of the expenditures for each regional and local government.

After the national Congress passes the public sector Budget Law (no later than 30 November), each subnational government must approve its opening institutional budget, which provides further details on the breakdown of subnational government expenditures. Subnational governments’ opening institutional budgets must be approved by their assemblies by 31 December, which leaves very little time for discussions. In France, for example, the law stipulates that for all subnational governments with more than 3 500 inhabitants, a debate on the main orientations of the budget must be organised at least 8 weeks before the vote for the regions and 10 weeks before the vote for the municipalities (Box ‎2.16).

During the budget formulation process, the MEF and the budget units (including subnational governments) discuss the budget units "revenue estimates". These revenue estimates do not only refer to own revenues, but are a discussion between the MEF and the budget units on how regular resources are to be allocated among budget units. This is generally referred to as a discussion about the level of expenditures of the budget units (sub-national governments) and thus of transfers between the national and sub-national governments. In this case, the allocation of regular resources between government entities results from political discussions, and is not governed by a formula or process that guarantees equity, predictability and transparency.

The structure of subnational government opening institutional budgets is similar to that of the public sector; in particular, subnational government budgets also link expenditures to funding sources, and revenues are not presented by revenue source (types of taxes, fees, etc.). In the budgets of OECD countries, for both national and subnational governments, there are two distinct sections: a section indicating the estimated revenue of the entity (national or subnational) and a section showing how these resources will be spent (Box ‎2.14).

The Budget Law applies to subnational governments and has the advantage of ensuring the homogeneity of subnational government budget processes, but this decreases flexibility to tailor the process to the size and type of subnational government. In many countries, subnational budget formulation and approval processes vary across subnational governments (especially between states in federal countries). France, for example, undertook a far-reaching reform of its public budget system in 2001, implementing performance budgeting. However, the new Organic Budget Law does not apply to subnational governments. Many subnational governments have been inspired by the new national Organic Budget Law, but each has done so in its own way and at its own pace.

Taxes collected by subnational governments represent less than 5% of subnational governments’ revenues in Peru (WOFI, 2022[26]). The average for unitary OECD countries is 34%; it is 37% for federal and quasi-federal countries. Transfers (most of ordinary revenue, donations and transfers, determined revenue) represent almost 90% of subnational government revenues in Peru, while they represent only 52% in unitary OECD countries and 46% in federal countries (Figure ‎2.14).

The amounts of FONCOMUN/FONCOR and mining canon that correspond to each sub-national government is determined by a formula, and the resources directly collected depend on each entity. Other public resources (in particular ordinary revenues) are considered as public sector revenues, and are not allocated to a particular level of government following any predetermined formula or process.

The composition of regional and local government funding sources shows that unrestricted transfers (directly collected revenue + determined revenue) represent, on average, 8% of regional government funding sources but 70% for local governments (Figure ‎2.15).

As with the national government, the link between expenditures and funding sources makes the budget formulation process particularly complex. Many OECD countries prohibit earmarking funding sources for specific expenditures for both national and subnational governments (Box ‎2.15). In most countries, the “no earmarking rule” prohibits tying funding sources to specific expenditures. For example, many countries, such as France and Spain (Box ‎2.15), also have a “no contraction rule”, which makes it mandatory to record all public sector revenues and expenditures, even if they balance. Therefore, the allocation of a revenue of national nature (such as the ordinary revenue in Peru) to an expenditure in a subnational government should be recorded, according to this principle, as an expenditure for the national government and as a transfer revenue for the subnational government. Accounting then consolidates these two transactions so that there is no double counting of the same expense.

OECD countries present the different sources of revenue in their budgets, but there is no one-to-one link between a specific revenue source and a specific expenditure; rather, the sum of revenues plus authorised deficits must cover the sum of expenditures. Most countries present revenues and expenditures classified according to economic criteria, distinguishing current, capital (or investment) and financial expenditures and revenues. This classification also makes it possible to calculate the balances for current operations (gross savings), the balance of the non-financial budget, and the budget’s borrowing capacity or need. This is also the classification most consistent with fiscal statistics; the non-financial budget balance, after adjustments, provides information on the net lending or borrowing, i.e. the deficit or surplus. In addition, the level of total government debt, and the amount to be issued if any, can be known (Figure ‎2.2).

In OECD countries, the concept that enables or authorises spending units to spend is the budget appropriation, spending appropriation or appropriation. Expenditure appropriations become spending appropriations once the budget bill is passed into law by the legislature. This is the concept that starts the chain of public expenditure, without the units to which the appropriation is attributed having to pay attention to how the particular expenditure is financed. As a general rule, spending appropriations are subject to three constraints: qualitative, quantitative and temporal, as they identify the destination of the expenditure, the amount and the year (calendar or fiscal) in which it is to be executed. Exceptionally, countries identify their own rules for modifying these appropriations, which are regulated in the organic budget law.

During the budget formulation process, as for other budget units, the expenditure “ceilings” communicated at the beginning of the budget process are considered by all as a starting point for discussions, which will be increased. Subnational governments do not thus know the exact amount that will be allocated to them when planning and prioritising their expenditures for the coming year.

In addition, the opening institutional budget does not include the amounts of FONCOR and FONCOMUN (as they are only allocated after the regularisation of income tax in March/April of the year of execution), nor does it include the total carry-overs (as the exact amount is only known in March) or the mining royalties. Before 2020, mining royalties were only included when they were realised, i.e. in June/July of the year of execution. Since 2020, regional and local governments benefiting from mining royalties have been allowed to incorporate 50% of the estimates from the opening institutional budget (DL 1441). The opening institutional budget also does not include rewards for meeting targets, which are also made late in the year and are immediately incorporated into the budget.

The fact that the opening institutional budget does not include all of the resources that subnational governments will have available does not allow the budget to be an efficient instrument for prioritising and planning expenditures, which reduced the quality of spending.

In all countries, the late determination of subnational transfers is a challenge for subnational government budgeting. However, no OECD country has budget modifications as significant as Peru. Several countries have developed processes to ensure that the budget is a credible and realistic picture of what will be implemented. For example, Spain has a “two-year liquidation system”: the subnational government budget indicates “revenues on account” (on account of the revenues expected to be collected during the year), and after two years a “liquidation” is carried out to correct the accounts according to the revenues actually collected. If these are higher than the “revenues on account”, a balance is generated in favour of the subnational government. If these are lower, the subnational government must repay the money unduly collected. France has a system where subnational government budgets can be voted on up to 15 April of the budget year, with rules so that recurrent and investment expenditure can still be met (Box ‎2.16).

It is common for federal or decentralised OECD countries to have formalised co-ordination mechanisms so that the subnational governments are informed by the national government of the share of revenue or revenue transfers that they will be allocated. For example, in the case of Spain, the Organic Law on Financing of the Autonomous Communities (regions) created the Fiscal and Financial Policy Council in 1980 to adapt the co-ordination between the financial activity of the autonomous communities and the State Treasury.

As discussed in Section ‎2.6, Figure ‎2.6, budget modifications are particularly important for subnational governments. The opening institutional budget is multiplied by 2 in regional governments and by eight in local governments. In such conditions, the budget cannot play its role in prioritising and planning expenditures.

All OECD countries have mechanisms to amend budgets during the year. However, they tend to accumulate several amendments and make amendments once or twice a year, when the amendments are significant or change expenditure ceilings and require legal approval (indeed, expenditure managers often have autonomy for internal reclassifications, which do not require legal approval). The changes made are also usually relatively small. For example, the additional budget of the Île-de-France Region in 2022 represented an increase of less than 1% of current revenues and expenditures and less than 10% of investment revenues and expenditures (despite the fact that the economic situation in 2022 was complex, with high inflation and lower than expected economic growth). Indeed, in the face of a 5.6% decrease in expected revenues, the region used part of its reserves to maintain its level of expenditure (and still comply with the balance rule).

Subnational governments’ strategic planning of revenues and expenditures within fiscally sustainable limits requires a high level of capacity. The OECD countries where subnational governments have the largest responsibilities (both as a percentage of total expenditure, revenue and expenditure autonomy, borrowing capacity, etc.) are also those where the overall capacity level of the public administration is the highest. In most OECD countries, subnational governments have difficulties attracting and retaining qualified staff. Indeed, subnational governments often pay lower salaries than the national government, so it is common for the most efficient employees, once trained, to move to work for the national government. Another frequent challenge is that the size of subnational governments would not require high-level capacity staff in large numbers (perhaps part-time would be sufficient) and does not offer staff attractive career prospects. To overcome this challenge, several countries are trying to create partnerships of municipalities for services (shared services), to attract qualified staff by offering higher salaries and sharing these resources among several subnational governments (Box ‎2.18).

Peru has the same challenge as most countries, with difficulties in attracting and retaining the skilled staff subnational governments need to manage their budgets strategically and sustainably.

Table ‎2.16 presents the OECD’s recommendations based on the analysis presented in the previous sections.

In 2007, Peru introduced “budget programmes” in its budget, according to a methodology developed by the MEF as part of the reform of the national budget system. This reform aimed to improve the effectiveness of public spending by aligning expenditures with government priorities and linking them to specific objectives. At the time, this reform constituted significant progress compared to previous practices, where resources were allocated by historical file without a clear assessment of needs or how to achieve the desired results.

In 2008, the budget comprised five budget programmes. The Articulated Nutrition Programme is one of the best known (Box ‎2.19). In the first years, the logic model was used to develop the programmes, which allowed taking into account actions between different ministries. However, the implementation of these programmes proved to be very complex and the logical framework, which is more operational but lacks the inter-sectoral dimension, was adopted. Today, the budget includes 88 budgetary programmes (of which 2 are multisectoral), representing 63% of public spending.8 The remaining resources are classified as budget allocations not linked to outputs or core activities (related to the management of equipment, human and financial resources).

The allocation of resources to each budget programme during the budget formulation process is carried out using a very micro approach, starting from individuals to global aggregates. At the beginning of the process, each executing unit plans its needs in terms of budget programmes and extrabudgetary programmes, which it submits to the MEF. The budget programmes distinguish key interventions, activities and management. The budget for key interventions is defined according to a combination of inputs identified during the budget programme formulation process, which is not regularly reviewed. Implementing units use information on the characteristics of their population (e.g. for health programmes), regional priorities and coverage targets, and determine their “physical targets”. These physical targets are entered into the system, which determines the quantities of each input needed at a very micro level of detail. The human resource needs are added to these combinations of inputs, as defined by the relevant directives. In health, for example, the Ministry of Health determines the number of nurses needed according to the expected number of patients, the characteristics of the health centre, etc.). The needs of each implementing unit are aggregated at the regional level and transmitted to the MEF, which allocates the necessary funds in the budget (Vammalle et al., 2018[30]).

This focus on results in the budget process is an important step forward, as it introduces the notion of evidence-based policy making, with results objectives and monitoring of targets. In particular, the identification of government priorities is interesting. However, the way in which these budgetary programmes are managed and included in the budget is done at a very micro and rigid level, with very little autonomy for ministries and budget units to manage their funds. This methodology allows identifying the efficient mix of inputs to achieve results and somewhat replaces the “input budget” with an “input mix budget”. But the budget is still input-based, with significant rigidities in moving funds from one line to another and adapting to fluctuating situations and challenges.

Most countries are moving away from budget practices based on detailed line items and inputs to develop budget processes that give greater autonomy to programme managers, executing units and line ministries. These reforms go hand-in-hand with efforts to strengthen line ministries’ capacities and create the fiscal resource management directorates presented in Chapter 4. There are different practices for integrating results information into budgets and moving from input-based budgets to programme-based budgets (Box ‎2.20).

To design a budget programme, we seek to use the best available evidence to identify causes and effects around the condition of interest, as well as the most efficient ways to achieve the desired results. To do this, we start by defining the “condition of interest”, i.e. the problem or situation that we want to influence because it is of interest for sustainable development. Next, an explanatory model is created, which identifies and prioritises the causes and factors of the condition of interest, based on evidence. The next stage is to develop a prescriptive model, or set of evidence-based interventions that respond to the explanatory model. Finally, the Theory of Change is used and a graphical summary of interventions and outputs is produced to achieve outcomes.

This describes good practice for developing any public policy, widely used in OECD countries. It corresponds, for example, to the methodology developed in England in the Green Book: Central Government Guidance on Appraisal and Evaluation (HM Treasury, 2022[32]). However, while OECD countries use this methodology to develop public policies, they do not use it to base budget allocations.

Indeed, basing budget allocations on the mix of actions and inputs identified in this methodology implies three fundamental implicit assumptions: 1) that the optimal mix of inputs is identical in all regions; 2) that it is identical whatever the scale of the intervention; and 3) that it does not vary over time. However, each of these assumptions is highly implausible; take the Articulated Nutritional Programme, for example. The main causes of undernutrition in children may differ between urban and rural, dry and wet regions, etc. Some investments may be cost-effective when developing a large-scale project but may not be justified for smaller project scales. Finally, the relative technologies and costs of different inputs vary over time so that the optimal mix at one point in time may not be optimal a few years later.

This is why in OECD countries that use programme budgets, the targets to be achieved by each programme manager and the overall pool of resources allocated to them are determined, but it is left to the discretion of each programme manager to implement their own logic model and determine the optimal mix of expenditures, depending on the targets and the resources available to them.

As mentioned above, this requires a high degree of administrative capacity in line ministries, executing units or subnational governments, and strong financial management directorates in ministries. Designing and developing a performance budgeting system is a long and difficult task, and in most countries is still an ongoing process. In particular, it is important to balance the efforts and costs involved with the benefits to be reaped. The OECD developed a guide to implementing performance budgeting, proposing good practices illustrated by concrete examples (Box ‎2.21). The OECD also regularly accompanies countries in designing their performance budgeting system and transitioning from the existing system to the new one.

Peru uses tools that grant additional resources to the budget units to help public entities achieve the objectives and results prioritised by the national government to improve the provision of public services. These incentives can be targeted at any actor in the results chain (budget units, regional or local governments). There are three types of programmes: the Budget Support Agreement, the Incentive Programme for the Improvement of Municipal Management and the Recognition of the Execution of Investments.

No OECD country uses a direct relationship between the achievement of results and resources allocated (see Box ‎2.20 and Figure ‎2.17).

Based on the analysis presented in the previous sections, this report proposes the following recommendations for improving the link between the expenditure budget and government priorities (Table ‎2.17).

Peru was a pioneer in the region in terms of initiating the transition to accrual accounting based on the International Public Sector Accounting Standards (IPSAS). Peru translated IPSAS into national standards and made key adjustments to move to accrual accounting. This provides a solid basis for agreeing on the content of the financial statements and for productive work between the different institutions. Preparing complete financial statements based on international standards represents a good practice set out in the OECD recommendations. The General Directorate of Public Accounting is in the process of fully adopting IPSAS, which leads to the application of international standards in recording the entities’ economic events (accrual accounting). To this end, it is important to establish the baseline of financial information through an ongoing process of accounting cleansing and reconciliation.

Another important issue to study to allow issuing financial statements based on international standards is related to revising the regulation of the Public Sector Financial Administration to improve the definition of the entities’ managers’ responsibilities.

The published accounts (the General Account of the Republic) consolidate the entire public sector. They are published annually and sent to Congress. They are accompanied by an opinion from the supreme audit institution, the Comptroller General of the Republic (CGR). One of the strengths of the system is the publication of consolidated financial statements, in accordance with known standards and subject to external audit. Congress approves a budget covering the entire public sector. Therefore, the General Account of the Republic covers all public sector institutions. The comprehensive character of the account increases its relevance. However, it also means that its accuracy depends on the quality of the work of numerous agents in many agencies. Providing these staff with sufficient and relevant training is a challenge. Ensuring that all major public sector bodies are subject to a high quality, constructive and independent annual audit would increase the assurance of the overall account. In this regard, it would be useful to assess the capacity of the audit firms to audit all entities under the scope of the National Accounting System, the technical level of the professionals charged with carrying out the review of the financial statements, and the capacity of the CGR to oversee the work carried out by the audit firms.

The General Account is a useful document that enhances accountability and supports congressional scrutiny of public finances. Ongoing work to improve the accounts by revamping standards, training staff and creating a body of professional accountants within the public sector has the potential to add value. This work must take into account the views of the CGR. Significant emphasis should be given to the professional training of agents with an accounting function.

Over the years, the CGR has issued a qualified opinion and more recently (COVID restrictions being a key factor) refrained from issuing an opinion. While it is common (e.g. in the European Union, France and the United Kingdom) for ambitious new accounting frameworks to require some years before all required information reaches a good level of assurance, this limited assurance from the external auditor reduces the confidence in public finances. The CGR has contributed to congressional decisions not to approve the General Account over the last eight years. While Congress has not approved the accounts in recent years, it has made use of them to provide a compelling document on the financial situation of the Peruvian public sector.9 It is important to establish and maintain a dialogue between the auditor and the comptroller in order to resolve, over time, the various observations on which their opinion is based. This would thus reduce the risk of non-approval by the Congressional Review Commission and give greater credibility to the quality of the information.

The accounting transactions of the entire Peruvian public sector are recorded in the Financial Administration System (SIAF). Although the diagnosis and analysis in the framework of this project did not specifically cover the functioning of the system, it is understood that it does not provide real-time access to accounting information, even though, in principle, it is adapted for treasury and budget operations. A key operational problem has been the absence of regular bank reconciliations in some entities, which has contributed to qualified audit opinions. See more on this issue in Chapter 3. To generate timely, quality and decision-useful information, it is not only necessary to achieve full adoption of IPSAS, but also to implement online accounting, which will be achieved to a large extent through the implementation of an integrated IT tool that allows economic events to be recorded on line. In this regard, an institutional project is being developed that consists of designing, developing and implementing the SIAF-RP, where it will be essential to include the necessary guidelines that allow international standards to be applied and generate the information that facilitates entities’ management and decision making.

The Comptroller General’s Office has highlighted incomplete information from taxation (in relation to its access to individual taxpayers’ records) and incomplete information from the Ministry of Transport and Communications (in particular on assets created through concessions). There are a number of cases around the world where supreme audit institutions have limitations in accessing taxpayer records. Many have identified mechanisms to avoid qualifying their opinion on the accounts as a result. Accounting for infrastructure assets needs to be consistent across the public sector.

The MEF’s General Directorate of Public Accounting prepares the accounting standards and consolidates the accounting statements of the various Peruvian public sector bodies to produce the final account. The CGR is responsible for the audit. In practice, auditing is usually carried out by private auditing firms and does not cover all agencies in a given year. Therefore, a key factor for the successful production of the accounts is the provision of detailed guidance and training. The MEF is specifically considering mechanisms to create a corps of trained and certified financial managers (in line with proposals in some other areas).

The Comptroller General’s Office has the role of external auditor for the Peruvian public sector. It also oversees internal audit. The CGR places a great emphasis on the role of “concurrent audit”, on its anti-corruption function and, in general, on compliance auditing.

The approach to concurrent audit is not in line with international best practice (which tends to focus on independent and ex post audit). There seems to be a conflict between measures to introduce concurrent audits and measures to simplify procedures for projects (see Chapter 5). Problems of corruption and non-completion of projects may make the early involvement of an independent outsider seem attractive. However, this practice also creates the risk that completed projects are recorded as incomplete for a long time after the works have been completed and the supplier has been paid. In this sense, an advisory function, rather than an audit requirement, might be a better suited to achieve this.

A paper published by the CGR suggests that delivery times increase by an average of six weeks for projects subject to concurrent audits. Reducing the time projects remain open would improve the control environment, and measures to accelerate all phases of project implementation would improve delivery.

As noted in the section on the accounts, the CGR’s work on the General Account results in a small number of issues leading to an appraisal of the accounts. Further bilateral contacts between the Comptroller and the MEF should identify ways to reduce the number of areas subject to qualification. This could involve, for example, clearer statements on the scope of consolidation of the General Account and on the audit opinion.

Based on the analysis presented in the previous sections, Table ‎2.18 presents the recommendations for improving the accounting and control of public accounts.

References

[33] Contraloría General de la República (2022), Evaluación de la Credibilidad presupuestal del gasto público en el Perú, https://cdn.www.gob.pe/uploads/document/file/3284646/Evaluaci%C3%B3n%20de%20la%20credibilidad%20presupuestal%20del%20ga.pdf (accessed on 26 November 2022).

[9] Department of Public Expenditure and Reform (2019), Public Financial Procedures Booklet, Department of Public Expenditures and Reform, https://www.gov.ie/en/publication/5268bd-government-accounting (accessed on 26 November 2022).

[2] Downes, R. et al. (2017), “Strengthening budget institutions in Public Expenditure Management Peer Assisted Learning (PEMPAL) countries: Results of the 2013-14 OECD PEMPAL Budget Practices and Procedures Survey”, OECD Journal on Budgeting, https://doi.org/10.1787/budget-16-5jfq810f35vc.

[23] French Ministry of Economy, Finances, and Industrial and Digital Sovereignty (2022), Website of the French Ministry of Economy, Finances, and Industrial and Digital Sovereignty, https://www.budget.gouv.fr (accessed on 26 November 2022).

[28] Government of France (2022), Code général des collectivités territoriales, https://www.legifrance.gouv.fr/codes/id/LEGITEXT000006070633 (accessed on 14 November 2022).

[32] HM Treasury (2022), Green Book: Central Government Guidance on Appraisal and Evaluation, UK Government, London, https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-governent.

[25] Kantorowicz, J. and B. Van Grieken (2019), 5th Annual Meeting of the Network on Fiscal Relations Across Levels of Government Spending Autonomy of Sub-central Governments: Conceptualisation and Measurement, background paper, https://www.oecd.org/tax/federalism/spending-autonomy-of-sub-central-governments-conceptualisation-and-measurement-background-paper.pdf.

[12] MEF (2022), Plataforma digital única del Estado Peruano, Ministry of Economy and Finance, Lima, https://www.gob.pe/mef (accessed on 26 November 2022).

[19] MEF (2021), Informe Global de la Gestión Presupuestaria 2019, Ministry of Economy and Finance, Lima, https://www.mef.gob.pe/contenidos/presu_publ/presu_sect/Evaluacion_Global_2019.pdf.

[21] MEF (2021), Informe Global de la Gestión Presupuestaria 2020, Ministry of Economy and Finance, Lima, https://www.mef.gob.pe/contenidos/presu_publ/presu_sect/Evaluacion_Global_2020.pdf.

[5] MEF (2020), Clasificador de Fuentes de Financiamiento y Rubros para el Año Fiscal, Ministry of Economy and Finance, Lima.

[17] MEF (2019), Evaluación Global de la Gestión Presupuestaria 2018, Ministry of Economy and Finance, Lima, https://www.mef.gob.pe/contenidos/presu_publ/presu_sect/Evaluacion_Global_2018.pdf.

[18] MEF (2018), Evaluación Global de la Gestión Presupuestaria 2017, Ministry of Economy and Finance, Lima, https://www.mef.gob.pe/contenidos/presu_publ/presu_sect/Evaluacion_Global_2017.pdf.

[20] MEF (2017), Evaluación Global de la Gestión Presupuestaria 2016, Ministry of Economy and Finance, Lima, https://www.mef.gob.pe/contenidos/presu_publ/presu_sect/Evaluacion_Global_2016.pdf.

[3] Moretti, D. (2019), “Budgeting in Greece”, OECD Journal on Budgeting, Vol. 2019/2, https://doi.org/10.1787/2f5e7d7a-en.

[24] OECD (2021), Fiscal Federalism 2022: Making Decentralisation Work, OECD Publishing, Paris, https://doi.org/10.1787/201c75b6-en.

[29] OECD (2020), The Future of Regional Development and Public Investment in Wales, United Kingdom, OECD Multi-level Governance Studies, OECD Publishing, Paris, https://doi.org/10.1787/e6f5201d-en.

[14] OECD (2019), Budgeting and Public Expenditures in OECD Countries 2019, OECD Publishing, Paris, https://doi.org/10.1787/9789264307957-en.

[7] OECD (2019), Government at a Glance 2019, OECD Publishing, Paris, https://doi.org/10.1787/8ccf5c38-en.

[27] OECD (2019), OECD Fiscal Decentralisation Database, https://www.oecd.org/tax/federalism/fiscal-decentralisation-database (accessed on 14 November 2022).

[31] OECD (2019), OECD Good Practices for Performance Budgeting, OECD Publishing, Paris, https://doi.org/10.1787/c90b0305-en.

[15] OECD (2018), OECD Performance Budgeting Survey, OECD, Paris.

[1] OECD (2015), “Recommendation of the Council on Budgetary Governance”, OECD Legal Instruments, OECD/LEGAL/0410, OECD, Paris, https://www.oecd.org/gov/budgeting/Recommendation-of-the-Council-on-Budgetary-Governance.pdf.

[6] OECD (2015), Recommendation of the Council on Budgetary Governance, OECD, Paris, https://www.oecd.org/gov/budgeting/Recommendation-of-the-Council-on-Budgetary-Governance.pdf.

[8] OECD (2014), Budgeting Practices and Procedures in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/9789264059696-en.

[13] OECD (forthcoming), OECD Spending Better Framework.

[22] OMB (2018), “Section 120 – Apportionment process”, OMB Circular No. A–11 (2022), Office of Management and Budget, Washington, DC, https://www.whitehouse.gov/wp-content/uploads/2018/06/s120.pdf.

[11] Ortiz, M. and D. Winkelried (eds.) (2022), Hitos de la Reforma Macroeconómica en el Perú 1990-2020: La Recompensa de los Tamías, Universidad del Pacífico, https://repositorio.up.edu.pe/bitstream/handle/11354/3337/OrtizWinkelried2022Cap9.pdf (accessed on 27 November 2022).

[10] Schmidt-Hebbel, K. (2022), “Política y reglas fiscales: evaluación y propuestas de reforma para el Perú”, http://www.schmidt-hebbel.com/assets/politica-y-reglas-fiscales----ksh-(1).pdf (accessed on 27 November 2022).

[16] Shack, N. and R. Rivera (2022), “Evaluación de la credibilidad presupuestal del gasto público en el Perú”, Documento de Política en Control Gubernamental, Office of the Comptroller General, Lima, https://cdn.www.gob.pe/uploads/document/file/3284646/Evaluaci%C3%B3n%20de%20la%20credibilidad%20presupuestal%20del%20ga.pdf.

[4] Vammalle, C. and I. Bambalaite (2021), “Fiscal rules for subnational governments: The devil’s in the details”, OECD Working Papers on Fiscal Federalism, No. 35, OECD Publishing, Paris, https://doi.org/10.1787/531da6f9-en.

[30] Vammalle, C. et al. (2018), “Financing and budgeting practices for health in Peru”, OECD Journal on Budgeting, https://doi.org/10.1787/budget-17-5j8v16g3czth.

[26] WOFI (2022), World Observatory on Subnational Government Finance and Investment website, https://www.sng-wofi.org (accessed on 14 November 2022).

Notes

← 1. It was decreed by Law No. 30823, which grants power to the executive branch to legislate in terms of economic management and paragraph a.2) of Section 5 of Article 2 of Law No. 30823, which establishes that the executive branch has the power to legislate on the modernisation of the budget, following Article 104 of the Constitution, which grants the executive branch the power to legislate through legislative decrees.

← 2. A budget line is the most detailed level of expenditure voted in the Budget Law. It can refer to a particular input (e.g. salaries) or be more aggregated to programme or budgeting unit level.

← 3. The calculation of some transfers takes into account the revenues or revenue-generating capacity of subnational governments, but the outcome of present collections will influence the formula and future transfers, not those of the current year.

← 4. The non-financial public sector is defined as all non-financial public sector entities, i.e. general government sector and non-financial public enterprises.

← 5. The values shown for the national government correspond to the balance of all positive and negative changes in absolute value. The data available did not allow for the sum of all changes, positive and negative, and therefore correspond to an underestimation of the size of the changes.

← 6. The CRG (2022[33]) study comes to the same conclusion.

← 7. Legislative Decree No. 1440, Article 37.1.

← 8. Data for 2022 provided by the MEF.

← 9. Opinion of the Budget and Accounting Committee of the Republic, Lima, 13 October 2021.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2023

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.