I. Building the Forward Estimates
1. Following the recommendations of the February 2017 Fiscal Affairs Department (FAD) mission, the Ministry of Finance (MoF) has begun preparing forward estimates as part of the broader medium-term budget reforms. This section provides an overview of the preparation of forward estimates in Poland and an update on progress since the February 2017 mission. The mission worked closely with the budget departments1 and line ministries to apply the forward estimates to individual ministries, and identified specific as well as cross cutting issues and recommendations on how to address them.
2. Forward estimates provide a starting point for setting expenditure allocations over the medium term. Forward estimates are bottom-up expenditure forecasts for each agency based on a no-policy change assumption, that is, in the absence of any new or amended laws, decisions by the Council of Ministers (CoM), or approvals by the MoF. As they are rolled over from one vintage to another, they incorporate and separately identify the impact of new decisions as they are made and presented in the updates.
3. Forward estimates represent neither the budget bids of agencies nor the actual spending allocations. They provide an indication on the cost of ongoing policies in future years, and are usually prepared early in the budget process. Budget bids, which are generally prepared at a later stage, consist of the assessment of the needs of each ministry, and may include new policy decisions that are expected to be taken. Eventually, actual spending allocations will not necessarily be set in line with the forward estimates—they will vary as new policy decisions are taken and as the impact of revisions to underlying economic parameters is accounted for.
4. The preparation of forward estimates enhances the budgetary process in several ways by:
Simplifying budget negotiations by shifting the focus away from discussing the budget item-by-item to developing new policies and evaluating the quality of existing policies;
Enabling the assessment of the fiscal space available in the medium term. This helps policy makers to decide now on which policies to adopt for future years. This assessment is relevant both when there is space for new policies or when savings are required to fit spending within the resource envelope; and
Allowing the early identification of pressures in individual spending areas, as well as the sources of these pressures. This is particularly important when corrective measures take time to come into effect.
5. The preparation of forward estimates at the disaggregated level includes a number of key steps. These are focused on understanding the existing budget and applying the medium-term cost drivers, and aggregating the forward estimates. Box 1 details the major steps in a commonly-used approach, that can be applied across ministries. The reliability of the forward estimates relies upon the consistent application of the methodology across each of the spending areas.
Box 1.Major Steps in Preparing Forward Estimates
1. Understand the existing budget
a. Understand the current spending base. This requires knowing where and what the agency spends on.
b. Choose the level on which to forecast spending. This does not need to be based on the lowest level of budgetary appropriation. Sections or paragraphs can be grouped so long as they have common cost drivers.
c. Identify past one-off expenditures. These are to be removed from the spending base if they will not be incurred in coming years (e.g., the cost of elections held in the base year).
2. Understand and apply medium term cost drivers
d. Identify price and volume cost drivers. This should be done for each level of forward estimates.
e. Link these cost drivers to macroeconomic and demographic variables (e.g., Consumer Price Index (CPI) or population growth).
f. Adjust base spending by the price and volume parameters
3. Include the effect of past policy decisions that are not fully reflected in the base
g. The additional costs or savings resulting from past policy decisions that are yet to fully mature are to be included to the adjusted projected spending (e.g., a policy of increasing transfers to a given group which was only implemented half way through the base year, or an IT modernization program that has not been fully implemented). Note that these costs/savings are expected to be already adjusted for price and volume parameters.
4. Aggregate to get the overall forward estimate for each Ministry
h. The set of forward estimates should be aggregated for each Ministry.
i. The transfers from budgetary parts to other entities within the scope of the forward estimates being prepared (e.g., special purpose funds, executive agencies and other state legal persons) should be netted out to provide a consolidated forward estimate for each ministry.
No policy change
6. The no policy change assumption is a fundamental feature of preparing forward estimates. These estimates are intended to represent the best estimate of the cost of continuing existing polices at the currently agreed levels of service, while incorporating the impact of existing spending decisions that have been taken but not yet incorporated in the base.2
7. While for the majority of spending this may be clear, there will be difficult areas where the future level of spending is not explicit in existing policies. For this reason, a clear definition of no-policy change is required, and should be broadly understood by policymakers, budget departments and line ministries. A possible definition could be:
The level of spending that will continue to occur over the next three years in the absence of any new or amended laws, decisions by the CoM, or by the MoF.
8. In cases where the policy position is not clear, the default presumption should be the continuation of the current level of services. In most cases there will be a clear delineation, however there are a number of cases where further clarification may be required. For instance, while legislation may define a given level of service, this may not be currently met in the budget. In these cases, the current budget funding level should be viewed as the best indicator of the current policy of the government.
9. The EC has also provided guidance to member states on the no policy change assumption for the high-level forecasts under the Convergence Program that should also be considered. It sets out that the estimates:
“should include all interventions by the government to change past policy orientations that is specified in sufficient detail, as well as adopted or at least credibly announced, and has a direct incremental budgetary impact compared to the baseline.” The EU has also outlined a set of methodological principles to clarify what is, and what is not, compatible with the no policy change assumption.
Scope of the forward estimates
10. The forward estimates should be as comprehensive as possible to provide the necessary information for budget decision-making. The coverage of institutions should include all entities that have a direct or indirect impact on expenditure that is subject to the key fiscal target (see Box 2). In the case of Poland, this applies to all entities and funds that fall within the Stabilizing Expenditure Rule (SER). In addition, coverage should be extended to entities that do not fall within the SER, but whose spending has budgetary implications.
11. When preparing estimates of the transfers to entities outside of the State Budget, it may be necessary to forecast the overall financial position of these entities. This applies to agencies who receive top-up transfers that depend on the entities’ funding needs (expenditure minus revenue). An example is the Social Insurance Fund (FUS) which relies on State Budget transfers to make up the shortfall between its mandatory spending and its revenue from social security contributions. In this case, both contributions and payments will need to be forecast (as they currently already are). For transfers that are fixed or formula-driven, it is not necessary to forecast the overall financial position, as the transfers can be forecast independently, without need to understand the external entities’ total activities. An example is the subsidies to the Polish State Railways (PKP) for youth passes which depends on the number of users and not on the company’s financial position.
Comparison with other spending forecasts
12. There are already several medium-term fiscal forecasts being prepared, both at a top-down and bottom-up level. These forecasts are prepared to meet certain internal and European requirements, but the budget estimates, prepared at a later stage of budget planning, are not aligned with these forecasts. One of the main benefits of the forward estimates is that by incorporating them within the top-down fiscal forecasts, the convergence program will be prepared on a basis that is consistent with the annual budget. A broader discussion is provided in the February 2017 FAD TA report.3
13. The Macroeconomic Policy Department (MPD) provides top-down spending forecasts of the general government sector, that are included in the Convergence Program. These forecasts include detailed projections of major units of GG sector and top-down projections for the rest of the general government sector units based on broad macroeconomic parameters. Figure 1 compares the realization of these forecasts with the EU average. These forecasts in the past have proved to be relatively accurate (at least for the budget year), and meet the requirements of the Convergence Program. While these forecasts are comprehensive and extend beyond the State Budget (see Box 2 for a comparison of the three main definitions of the government sector in Poland), they ignore a large amount of information from line ministries that can be picked up within the bottom-up forward estimates.
Figure 1.Average Difference between Medium-Term Estimates and Actual Expenditure
Source: Convergence Programs 2004–16
Box 2.Sectoral Coverage in Poland
The General Government sector is the broadest measure of government in Poland, and is defined in accordance with ESA methodology. The European fiscal requirements apply to this measure. It includes, among others, nearly all entities covered by the Public Finance Sector (defined below), and some entities not covered by the Public Finance Sector, such as the National Road Fund. General government spending represented 41.2 percent of GDP in 2016. The Stabilizing Expenditure Rule (SER) covers around 90 percent of general government expenditures.
The Public Finance sector, which includes about 90 percent of the general government sector, covers entities enumerated in the Public Finance Act 2009. It includes the state budget entities, as well as local government entities, Special Purpose Funds, such as the Social Insurance Fund and the Labor Fund, some State Legal Persons, and the National Health Fund. The Special Purpose Funds are particularly important as they represent more than half of the spending of the Public Finance Sector. The largest of these is the Social Insurance Fund, which accounts for spending of PLN 217bn in 2017 (of which PLN 47bn was sourced from State Budget transfers).
The State Budget includes expenditure of state budget units (ministries, courts, tribunals, and other governmental bodies) and other expenditure not assigned to any state budget unit, such as the contribution to the EU. The Budget Act covers the State Budget, as well as the financial plans of Special Purpose Funds, Executive Agencies, Budget Economy Institutions and State Legal Persons.
14. Line ministries, other holders of budgetary parts, and certain types of extra-budgetary units provide their spending forecasts for the purpose of the Multiannual State Financial Plan. Until 2018 these forecasts were not used through the budget process, as they were produced solely by the line ministries and other budgetary part holders, without central guidance or verification, and have thus evolved to highly optimistic bids for future spending by the agencies, rather than realistic forecasts of future spending.
B. Progress of Forward Estimates
15. This year, a first attempt at preparing forward estimates was made and the available fiscal space was presented to the Minister of Finance and further discussed within the CoM. Traditionally, the annual budget process comprised of a single stage where line ministries submitted their bids that are discussed with the MoF, and policy decisions are made by the CoM at a late stage of the budget process. This year, forward estimates for 2018–20 were prepared by the Budget Departments of the MoF, and the resulting fiscal space was presented to the CoM in April 2017, prior to the adoption of the Multiannual State Financial Plan (which includes the Convergence Program), giving early warning in the budget cycle that based on current policy there are significant adjustments required to meet the SER.
16. Definition of a no-policy-change was prepared by the Expenditure Policy Department. Forward estimates were prepared by the Budget Departments on the basis of the same groupings used for the purpose of setting budget limits.
17. The forward estimates were compared to the current budget base to check for any major differences. This check did not identify any large deviations, and was straightforward to undertake as the task level of aggregation used for the forward estimates is also used for the annual budget process.
18. The forward estimates allow some critical comparisons to be made with other budget aggregates:
Stabilizing Expenditure Rule – to allow the identification of the medium-term fiscal space;
The MPD’s top-down expenditure forecasts – to assist in the validation of both sets of forecasts; and
Line ministries’ and other budget part holders’ multi-year spending forecasts – to identify the level of new policy bids implied by these forecasts.
19. Once the forward estimates mature, the next step will be for them to become an integral part of the budget process. These estimates should feed into the aggregate medium-term spending forecasts currently made by the MPD which so far use less detailed approaches. This will then be used for the Convergence Program and constitute the starting point for budget preparation.
C. Applying Forward Estimates to Select Line Ministries
20. The line ministries chosen for the application of the forward estimates approach during this mission span most of the key issues likely to be encountered in preparing forward estimates. The budget for the Ministry of Internal Affairs contains a significant proportion dedicated to the current operational expenditure of its subordinate units, with a large component of wages and salaries. The Ministry of Family, Labor and Social Policy’s portfolio (which includes related spending by voivods) is dominated by large demand driven entitlement programs, and the budget of the Ministry of Infrastructure and Construction is focused on investment spending.
21. The mission worked through the forward estimates with each line ministry, and their relevant budget department within the MoF. This took the form of day-long sessions where the conceptual approach, the methodology used by the Budget Departments in the March 2017 exercise, and the spending drivers of each ministry were discussed.
Ministry of Internal Affairs
22. The Ministry of Internal Affairs budget in 2017 is PLN 21.5bn (1.1 percent of GDP). The largest component of the budget of the Ministry is the operational expenditure of agencies providing public security, such as the police, and border guards, and fire prevention. Budget part 42 – Internal Affairs (see Box 3 for a description of the budget classification) comprises the major share of spending of this Ministry (Figure 2). In this part, current expenditure amounts to PLN 10.9bn, of which 80 percent is wages and salaries.4 The other major component of the budget of the Ministry is the payment of pensions to former members of the public security services (PLN 8.9bn).
Figure 2.2017 Budget of Ministry of Internal Affairs
Source: Ministry of Finance
Box 3.State Budget Division and Budget Classification in Poland
There are two systems of classification for the state budget. The main system, used by the Parliament to authorize and control expenditure, is based on 84 budgetary parts, which constitute the division of the state budget. These parts are held by line ministers, who can be holders of one or more budgetary parts, heads of central offices, voivods, and other units. Within parts, transactions are recorded in sections, which are further divided into chapters. Sections and chapters were intended to be broadly based on the functions and sub-functions of governments. However, in many cases, their application mixes functions, organizations, programs and projects. Paragraphs represent the economic dimension of classification, though these are not consistently applied and do not fully meet ESA/GFS needs. The Parliament authorizes and control expenditure at the level of budget parts, subdivided into sections and chapters and groups of paragraphs (grants, transfers to natural persons, current expenditures of budget units, investment, servicing the public debt, EU own resources, and co-financing of EU-funded projects).
In addition, alongside the traditional budget classification, a programmatic classification was introduced in 2009 which requires budget holders to develop their budget and multi-year spending forecasts along the lines of functions, tasks, subtasks, and activities. Performance information is presented mainly as an annex to the budget, but is not used in decision-making by budget holders or the MoF.
Informally, for setting budgetary limits, the MoF can use additional breakdowns, if they provide a better mapping to the main activities of the government. The March 2017 forward estimates were also prepared on this basis.
23. Other comparatively smaller items are transfers to outside bodies and minor capital spending. An example is the transfers to higher education institutions that train police and fire service officers. Investment spending on minor capital purchases accounts for 2 percent of the budget of the Ministry.
24. For current expenditure, the proposed approach is to assume no volume adjustment in the no-policy change baseline. While there is a case for volume adjustment in other areas of spending (e.g., education needs to be adjusted with the number of school-age children), services related to law and order are inherently discretionary in nature and changes in the volume of services are typically treated as policy decisions.
25. There are a small number of cases where the forward estimate should consider a change in the volume of activity. This can occur when the services are externally driven. An example is the number of border guards in the event there was a material change in migrant arrivals.
26. Within current expenditure, the adjustment for forecast price increases should be separated between wage and non-wage components. The current approach is to hold wages flat in nominal terms over the forward estimates. For non-wage costs, a broad measure of forecast inflation such as the consumer price index should be used, unless there is a strong case to move away from a standard measure of inflation (an example could be the cost of pharmaceuticals which typically rises faster than CPI). Generally, such a case cannot be made for this Ministry.
27. Current expenditure should be adjusted to include the effects of past policy decisions that are not fully reflected in the base. In this Ministry, an example is the policy decision to provide additional funds for the modernization of the police, which will require additional spending of PLN 0.47bn in 2018 and a further PLN 0.97bn in 2019. As this decision has already been taken by the CoM, but not fully reflected in the 2017 base, the forward estimates will need to be adjusted separately.
28. As is the case for other entitlement programs, pension payments to former officers are already being modelled, and they should form the basis of the forward estimates. As it is a mandatory program, the approach is to model the eligibility for the pension based on expected retirements, mortality and income levels and to adjust the estimates for the indexation of the payments. The indexation is set out in legislation and comprise of an inflation adjustment and an increase representing 20 percent of the growth in real wages in the economy in the previous year.
29. Parameters used in these models should be consistent with the macroeconomic forecasts. This ensures a consistent approach across ministries to the forecast of major benefit programs. If the approaches differ, the forecasts will be inconsistent, which can lead to inaccurate estimates in aggregate.
30. The treatment of the transfers in the forward estimates requires a case-by-case approach based on the cost and volume drivers of a no policy change scenario. In the case of transfers for the training of police officers, the volume adjustment may need to account for the forecast level of recruits needed to maintain current police force levels. This may also require a specific price index for higher education (consistent with other agencies that provide higher education, such as ministries of education, defense, and health) to account for the composition of spending within those organizations.
31. For the March 2017 exercise, the relevant Budget Department prepared forward estimates by broadly applying the appropriate model. While preparing the estimates, the key steps as outlined in Box 1 were followed.
32. There was broad acceptance by the Ministry of Internal Affairs for the proposed methodology. While they prepare their budget bids based on an assessment of additional needs, they understood the approach being put forward for estimates on a no-policy change basis.
33. Two outstanding issues relate to the treatment of wages and minor capital expenditure, which will be system-wide across the public sector. The proposed approach to the treatment of wages is to keep them flat in the forward estimates, in line with the approach adopted by MoF. Capital expenditure in this Ministry is for smaller capital purchases and the forward estimates should be based on rolling providing for the continuation of current capital expenditure levels in real terms.
34. Further consideration needs to be given to the level of disaggregation where budget parts comprise activities which could have significantly different cost drivers. As an example, police and border guards are classified under the same section, but these two areas of spending are likely to have different volume drivers and hence require separate estimation. Similar consideration around separation of spending areas when policy decisions impact these areas differently. In this case, the costs of modernization of the police force and of spending to meet the EU commitment to expand border management by the European Border and Coast Guard Agency (Frontex) will need to be factored in separately. It is still necessary to strike the right balance between separating activities with different cost drivers and the complication associated with further disaggregation.
35. The readily available grouping of paragraphs does not separate wage and non- wage spending. This does not facilitate the adjustment of forward estimates for current expenditure using separate indices. This issue should be addressed as part of the reforms of the CoA and Budget Classifications (see Chapter II). In the meantime, effort needs to be made to more easily identify wage and non-wage components within the current expenditure grouping of paragraphs.
Ministry of Family, Labor, and Social Policy
36. Spending within the portfolio of the Ministry is budgeted at PLN 265bn (14 percent of GDP) in 2017, consisting predominantly of the payments of mandatory legislated entitlements. Most of these payments, worth PLN 217bn, are made from the FUS, consisting of old age, disability and survivor pensions, and sickness and maternal allowances. The FUS receives transfers from the State Budget of PLN 47bn, and the rest of its payments is covered by social security contributions. The other main mandatory components of the Ministry’s portfolio are family and social benefits (including the new 500+ Family Program delivered through the voivods), and unemployment and pre-retirement benefits paid from the Labor Fund.
37. In addition, the Ministry funds active labor market and social support programs. There are active programs for the unemployed delivered through the Labor Fund (PLN 6.6bn) and for training and rehabilitating disabled persons delivered through the State Fund for the Rehabilitation of Disabled Persons (PLN 4.9bn). The budget parts also include spending for multiannual operational programs funded by the EU and spending by voivods
38. The demand driven entitlement programs are already modelled in detail either within this Ministry or by the Social Security Institution (ZUS) and should form the basis of the forward estimates. The modelling includes forecast numbers of beneficiaries based on demographic and macroeconomic assumptions, including incomes and labor markets. Price adjustments for the forward estimates are based on the type of indexation set out in the legislation. ZUS uses the macroeconomic forecasts prepared by MPD, which in turn makes use of its forecasts on social security spending to prepare the Convergence Program.
39. Frequent policy changes present a challenge to modeling. Examples of these on recent years include the changes to the retirement age and the introduction of the 500+ Family Program. It is important to identify the additional future costs of such policies that are not included in the base used for preparing forward estimates. More generally, the behavioral impact of these policies needs to be modelled.
40. Estimates for active programs that assist the unemployed and the disabled with gaining employment should also be based on volume and price forecasts. While these programs do not reflect legal entitlements, the proposed approach treats the current level of spending per unemployed or disabled person as the no policy change assumption. This allows the forward estimates to increase to meet additional demand but delivers savings when the number of beneficiaries falls. The volume forecasts are based on the projected number of relevant beneficiaries while the price forecasts depend on the nature of the program. For example, wage subsidies for disabled persons are a fixed amount; other programs would be adjusted for forecast inflation.
41. The approach to the components of State Budget funding should be based on the nature of the expenditure:
EU supported activities: estimates should reflect the agreed funding profile for those programs.
Spending by voivodships for services (e.g., social housing): estimates should be based on the current level of activity and adjusted for inflation and any one-offs.
Current expenditure: estimates should be based on the current level of activity and adjusted for inflation and any one-offs.
42. The forward estimates prepared by Budget Zone Financing Department are broadly consistent with the proposed approach. The estimates for demand driven programs are based on detailed estimates prepared by the line ministry which use existing methodologies to project recipient numbers and factor in economic parameters. The Ministry is well versed in preparing projections that would form the basis of forward estimates as they do this for their own planning purposes.
43. The size and complexity of the spending on entitlement programs are sound reasons to require detailed reporting on the methodology and assumptions behind forward estimates. This allows a broad review of methodologies and checks to be conducted on the consistency on the use of macroeconomic forecasts. This reporting should be based on templates provided centrally by the MoF and replaced where necessary by underlying data.
44. Growth projections should be linked to macroeconomic forecasts to the maximum extent possible. Unemployment benefits should be directly linked to the unemployment rate (while there is a difference between the number of unemployment benefit recipients and the those reported as unemployed, the change in those numbers should be highly correlated). Other working age payments such as the disability pension are indirectly related to the state of the labor market. Linking these projections for these programs to the available macroeconomic forecast ensures that a consistent judgment about the future labor market is being applied across the various programs.
Ministry of Infrastructure and Construction
45. General government capital spending was forecast to total PLN 81bn (4.2 percent of GDP) in 2017. Of this, 60 percent of the spending is by central government, and 40 percent by local government. Within the central government, PLN 20bn was from the State Budget (including subsidies to other units, mainly local governments units), PLN 14bn by the National Road Fund (NRF), with the remainder dominated by the railway infrastructure company (PKP PLK S.A.), R&D institutions and universities. Within the state budget, the largest components of capital spending are defense (around PLN 10bn) and, road and railway infrastructure (around PLN 3bn).
46. Capital spending under the responsibility of the Ministry accounts for PLN 21bn (1 percent of GDP) and covers around a quarter of general government capital spending. The direct capital spending by the Ministry through the state budget is around PLN3bn for road and railway infrastructure. It also has responsibility for the spending of both the Road and Rail Funds that sit outside the state budget.
47. Capital investment spending poses challenges for the preparation of forward estimates, but is also an area where the preparation of forward estimates yields the largest dividends. Large projects can span several years and depending on the profile of expenditure over time can result in significant changes in overall funding levels. There is also a question of what constitutes existing policy that should be included in the baseline, particularly for smaller projects or capital purchases where it is not practical or efficient for ministers to be involved in all decision making.
48. Reflecting current policy on capital investment in the forward estimates provides valuable information on the future cost of the policy and helps avoid over commitments. This information should include the future operational and maintenance costs of approved projects. This allows the CoM to make decisions on the profile of capital investments to fit within the fiscal target over the medium term. It requires the clear identification of the point in the project development cycle where individual projects are considered approved and to be included within the forward estimates as new policy decisions. If the decision point is too early in the cycle, such as inclusion in development plans, this may open the door for too many projects whose funding needs cannot be accommodated. If set too late in the cycle, such as when cash appropriations are included in the budget, the forward estimates will tend to understate capital expenditure.
49. The forward estimates for major capital investments should provide for approved projects but not allow estimates to be reallocated to other projects that are considered new policy. This is an important issue once forward estimates are used as the basis for budget allocations. The framework should ensure that forward estimates justified on the basis of approved projects are only available for those projects.
The following approach is proposed to prepare forward estimates for the major spending items identified above:
For road and rail infrastructure spending, use the estimates in the National Roads Construction Programme and National Railway Programme, as these represent firm policy approval by the CoM. Both programs include budgetary and extra-budgetary spending. The original cost estimates included in the programs need to be adjusted in the preparation of forward estimates for known price movements and changes in the timing of projects since the approval in 2015;
Subsidies for young rail users, being formula-based, can be forecast using projections for CPI and the targeted population groups provided by the Central Statistical Office as price and volume indicators respectively;
Transfers to the housing sector, which depends on the health of this sector, can be forecast using GDP growth, interest rates, and indicators of the financial conditions of households and banks;
Forward estimates for other current expenditures (administrative costs) can be based on allocations for the previous year adjusted for estimated price parameters. As wages are a significant component, it is important to identify their share in order to determine the appropriate price parameter; and
Maintenance costs should be adjusted for both price and volume changes. Price adjustments should be based on a forecast macroeconomic parameter such as a producer price index or the GDP deflator for construction. Volume changes reflect the addition of new infrastructure which will add to these costs.
50. The relevant budget department acknowledges that predicting capital spending for this Ministry requires a different approach from that used for other line ministries. Future capital outlays in road and rail infrastructure are given by their respective programs approved by the CoM, including the aggregate allocation to minor projects. In effect, they can be considered as the no-policy change estimates. The same is not true for other line ministries, where similar programs do not exist. A proposed approach for these line ministries was outlined in this section.
51. The accuracy of the forward estimates for this Ministry will depend on the availability of updated estimates of the cost of major projects. Since the National Road and Railway Programs were prepared in 2015 for the period 2015–23, there needs to be regular updating of the costs and time profile of spending to account for possible delays and cost changes. To accommodate changes in the price of construction projects (that tend to have far larger variations than CPI), one option would be to use a construction price index such as a producer price index or the GDP deflator for construction.
52. The additional maintenance costs required by new infrastructure are not typically factored in budgetary estimates. While it is true that the maintenance cost of new infrastructure tends to be small, particularly in the first year of operation, a multi-year perspective requires inclusion and consideration of these costs over the project’s lifecycle, and inclusion within the forward estimates.
D. Cross-cutting Issues
Wages and salaries
53. The approach used in the March 2017 exercise was to keep the level of nominal wages flat in the preparation of the forward estimates. After these forward estimates were aggregated, the implications of an overall increase in public sector wages by 1 percentage point was estimated. However, flat wages remained the working assumption behind the forward estimates.
54. Keeping nominal salaries flat over the longer term may not be a realistic assumption, particularly as labor market conditions are tightening. The unemployment rate in Poland is at a historic low. Wages in the budgetary sector have exhibited a positive real growth trend, and have closely tracked average wages in the economy (Figure 3). Keeping wages flat can thus lead to an understatement of expenditure in the forward estimates, which can result in an overstatement of the fiscal position.
Figure 3.Labor Market Pressures in Poland
Source: Central Statistics Office, Poland. Forecast of real wages are taken from the 2017 Convergence Program.
55. A possible approach is to include a provision for future wage increases in a contingency reserve. If this better reflects likely wage increases, it will improve the quality of the aggregate forward estimates. It also has the advantage of not prejudicing future salary negotiations with public sector unions, which could be the case if explicitly included within ministries’ own forward estimates. The provision could be calculated using a projected wage price index that is prepared as part of the macroeconomic parameters. Once a decision is made on the size of a wage increase, the required amounts would then be migrated to the individual line ministry forward estimates and a corresponding reduction in the contingency reserve made. Box 4 shows examples of how wages are treated in the forward estimates in a few countries.
Box 4.The Treatment of Wages in Forward Estimates – Country Examples
A key consideration in the treatment of wages in preparing forward estimates is to avoid prejudicing future wage negotiations, while keeping the estimates as realistic as possible. On one hand, a generous indexation of wages, to potentially account for the uncertainty around the outcomes of future wage negotiations, may create an expectation of high wage awards in the future. On the other hand, keeping wages flat in an environment where wages have been rising can lead to unrealistically low forward estimates. The examples below reflect how some countries have tried to address this issue.
Source: IMF missions
56. The general approach to price indexation should be to use the broad measure of inflation unless there is a compelling reason to do otherwise. A broad measure of inflation, such as the CPI, for which forecasts are readily available, would be suitable. The only exception should be when a budget holder had a major input that has price changes that are significantly different to the CPI. This should only be considered if it has a material impact on the budget of the organization. For example, the price of delivering health care or pharmaceuticals usually rises faster than CPI, while imported goods may require an exchange rate adjustment.
57. Calculation of the appropriate price adjustment for transfers, subsidies, and general spending on the provision of services pose specific challenges. Examples of these are transfers paid to higher education institutes and spending by voivods for social protection services. For these, the appropriate index will depend on the cost structure of the agency receiving the transfer, which will need to be ascertained. An additional challenge is that similar agencies may receive transfers from different budgetary parts. If the calculation of the appropriate method of price adjustment is left to a case by case approach, consistency may be difficult to achieve.
58. Provision of price indices to apply to particular sectors will ensure consistency across ministries. For example, calculation of a price index for transfers for higher education would ensure that the forward estimates for all such transfers were treated consistently. The cost structure of services delivered by voivods could be examined to areas of similarity and the best basis for price indexation.
59. Capital investment spending poses particular challenges for the preparation of forward estimates. Large projects can span several years and depending on the profile of expenditure over time can result in significant changes in overall funding levels. There is also a question of what constitutes existing policy that should be included in the baseline, particularly for minor projects or capital purchases where it is not practical or efficient for ministers to be involved in all decision making.
60. For major projects, the forward estimates should be based on the estimated cost over time of currently approved projects. Decisions on these projects would usually involve ministers and the extent of current policy should be clear. This is the approach proposed for the approved National Railway Program. The estimates for these large projects should take into account the expected impact of inflation over the life of the project. As projects move through the planning, feasibility, contracting, and delivery steps, the costs and timing of projects will change. The forward estimates need to be adjusted at each major update.
61. There should be a particular decision point where major projects are considered to be approved and hence included within the forward estimates. Projects move through a number of stages with development plans and feasibility studies before a decision is taken to begin construction. The crucial decision point is when the project is ultimately approved by the CoM. This should be approved on the basis of sound cash-flow requirements for the project, so that the budgetary implications of the decision are clear and understood. Vague, total project costs without specifying the time dimension will cause difficulties later on in the project cycle.
62. Updating of estimates for cost changes and new policy decisions is part of the process of preparing forward estimates. Vintages of forward estimates should be reconciled with identification of the main sources of variation.
63. For minor projects and capital purchases, the forward estimates should include an estimated level of expenditure based on current levels. If expenditure by ministries is reasonably stable over time, this is best calculated for each ministry based on current levels of spending. Where capital expenditure is variable over time, an average level of past expenditure should be included in the forward estimates, either at an aggregate level or for each ministry. If an amount is included in the forward estimates in aggregate, this amount could be subsequently allocated across ministries in the subsequent budget process. If the forward estimates are calculated for each ministry, a prescribed average period, such as five years, should be used to avoid different periods of averaging being used in different ministries. These historical averages need to be adjusted for inflation.
64. A threshold should be used to determine the size of capital investments that would be considered major projects and constitute new policy. The size of the threshold should take into account the large number, and relatively small size of projects usually undertaken across the state budget. Setting a threshold balances the need to scrutinize major projects as new policies, without impeding the undertaking of minor capital investments that form part of the ongoing activity of ministries.
E. Further Development of the Forward Estimates
Assuring the consistency of the forward estimates
65. Early attempts at the preparation of forward estimates are often not particularly accurate but checking for consistency is crucial for longer term success. This should be facilitated by providing ongoing guidance material and templates seeking adoption of a uniform approach in the early stages. Review of the forward estimates by a central unit within the MoF is also crucial to achieve consistency. This could include a top-down comparison with the forward estimates to see how they compare to the earlier forecasts prepared by the MPD.
66. Central review of the forward estimates would require access to the documentation on the methodology, assumptions and macroeconomic forecasts used. This is particularly important for large demand-driven programs with complex methodologies. Documentation should be based on a standard proforma, setting out the methodology and assumptions, and supplemented where necessary with more detailed calculations. There should be continual review of the validity of assumptions to identify errors in estimates forecasts.
Engagement with line ministries on the forward estimates
67. The quality of forward estimates over time will be improved by increasing engagement with line ministries on the method of their preparation. The first few attempts at the forward estimates should be undertaken jointly by the MoF and the line ministry as this allows greater guidance and checking in the early stages. Greater engagement with the line ministries over time allows them to improve the estimates based on their superior knowledge of their activities and their associated cost drivers. In a first stage, this could take the form of MoF consulting the line ministries when preparing the forward estimates. With line ministries acquiring the necessary skills, this could be followed by parallel preparation of forward estimates by both the Budget Departments and the line ministries to enable comparison and discussion.
Level on which to calculate forward estimates
68. The level of disaggregation in the forward estimates should strike a balance between complexity and tractability. The aim should be to break up expenditure into the level necessary to reflect different price and volume drivers of expenditure while avoiding undue complexity. This balance should also recognize that it is more important to get the estimation methods right for the largest components of the budget. As line ministries will be increasingly involved in the preparation of the forward estimates, it is important to get the balance right, to avoid adding complexity to their roles.
69. The current classification of the budget does not lend itself to budget prioritization and decision-making. The structure of budget parts, sections and chapters does not provide a clear picture of the funding provided to agencies and major areas of spending. For instance, Section 754 in budget part 42 combines the major spending areas of police and border security.
70. As a result, Budget Departments in certain cases have developed different groupings of government tasks. Forward estimates, in some cases, have been prepared following these groupings, which cut across parts, sections, and chapters to provide a better mapping to the major activities of government.
71. The development of a new system of budget classification should provide a consistent framework that can be used at each step of the budget process, from early planning to budget execution reporting. This will remove the need to use different approaches to the preparation of forward estimates (tasks) and the preparation of appropriations (see Chapter II).
F. Future work on Forward Estimates and the Medium-Term Budget Framework
72. To proceed with the medium-term budget reform, the MoF should follow the proposed time frames.
Engagement with remaining ministries on the methodology for the forward estimates.
Documentation of methodology of the forward estimates prepared by each budget department with clear assumptions, followed by an analysis of these methodologies to ensure consistency.
Budget departments to prepare forward estimates in consultation with line ministries and use forward estimates as a basis for budget discussion.
Compare the forward estimates to the top down fiscal forecasts prepare by MPD identify and reconcile major differences, and depending on confidence in the forward estimates, potentially use the forward estimates to inform the Convergence Program Fiscal forecasts.
Comparison of 2018 forward estimates to those prepared in 2017 and identification of drivers of major changes.
Budget departments and line ministries to both prepare forward estimates on the basis of agreed methodologies. Formally reconcile the reasons for variations between the 2018 and 2019 vintages of forward estimates.
In preparing the budget for 2020, the first estimates for the budget year should be based on a rollover of the forward estimates from March 2019.
Broader Medium-Term Budget Framework reforms
73. The forward estimates approach is the first tool of several tools that support an MTBF, which still requires further design choices. While not a focus of this mission (see the FAD March 2017 report for more details),5 these tools and design choices will be an important part of the MoF’s reforms, and should begin to be addressed according to the below timeframes:
Costing of new policy proposals (March 2018–June 2019). Medium-term cost estimates should be prepared for all new policy proposals, as well as changes to existing policies, including alterations to eligibility criteria or assistance rates of existing transfers or social security programs.
Spending reviews (ongoing). Poland has already put in place a framework of spending reviews, but their impact in identifying savings or reallocation options has been limited. To increase the effectiveness of the reviews, they should be given clear targets for achieving savings, be fully integrated with the budget process, and be staffed by representatives from the MoF’s budget departments, line ministries, and possibly external experts.
74. The design of the MTBF should be developed in order to meet the government’s fiscal objectives. Key aspects to be resolved are (i) the framework and basis of medium-term expenditure commitment; (ii) the multi-year prioritization process; (iii) the control mechanisms to enforce multi-annual spending commitments; (iv) the accountability mechanisms to establish credibility. The March 2017 FAD report makes recommendations that fit the Polish context in each of these areas and provides a time frame for the implementation of the reforms.
1.1 Develop a set of methodologies to ensure the consistency of the forward estimates:
Develop a set of price indices for the treatment of transfers;
Provide for the separate identification of wage and non-wage estimates within commonly used reporting; and
Require detailed reporting on the methodology, assumptions and macroeconomic forecasts used in the preparation of the forward estimates for major programs, with estimation methodologies linked to macroeconomic forecasts wherever possible;
1.2 Establish a process for the central validation of forward estimates within the MoF, which includes reviewing methodologies for consistency and analyzing major variations between vintages of forward estimates.
1.3 Assign responsibility for the preparation of the March 2018 Forward Estimates to the Budget Departments, working with line ministries under the central coordination of the State Budget Department.
1.4 Continue to leave nominal wage rates fixed, as per the CoM decision, but include a provision for future wage decisions within a contingency reserve.
1.5 Provide detailed guidance on the treatment of capital spending, differentiating between minor and major capital projects:
Provide a threshold to delineate between major and minor capital projects with major projects to constitute new policy;
Require that forward estimates for major projects be based on up-to-date costs and time profiles of agreed projects; and
Include the additional maintenance costs required by new infrastructure in forward estimates based on the estimates provided in the Regulatory Impact Assessment.
II. Developing a Standardized Chart of Accounts
75. A Chart of Accounts (CoA) underlies the appropriate classification, recording and reporting of the financial plans, activities, flows and stocks of government. A CoA presents a systematic and organized coded list of the individual accounts that are used to record flows of economic value between economic units and eventually report on the stock positions (balances) of assets and liabilities at a reporting date. A CoA should meet not only legal and administrative requirements for budget management, but also ensure that the necessary details are available to comply with international financial and statistical reporting standards and guidelines.
76. Standardizing the CoA for all government units of the general government sector improves the consistency of reporting and facilitates the aggregation and consolidation of the accounts of various units.6 Implementing a standard chart of accounts (SCoA) requires centralized control over the development and changes to the structure, classifications and interpretations of the code list.7 The SCoA is also set against a background of common accounting and reporting standards, principles and policies. However, this control is often limited to determining the higher levels of classification system, while some flexibility may be allowed at the lower levels of classification.
77. A SCoA can also be integrated with budget classifications. Such integration establish links between the budgetary accounts and the financial account used in the preparation of general purpose financial statements. These two different types of reports have different objectives, each following its own logic and presentational structures, and are produced for the benefit of different users. Despite these differences, an integrated account coding structure ensures consistency between budget and financial account presentations, to the extent possible. Where differences in objectives demand different aggregations or presentations, the integrated SCoA allows precise reconciliation.
78. The remainder of this section describes issues related to the plans to implement an integrated SCoA in Poland. The section describes the background to the implementation of a SCoA and then elaborates on the pre-conditions or assumptions underlying the implementation plan. The main findings of the mission are presented before the approach to the project, a proposed workplan and future follow-up needs are discussed.
79. Introducing a SCoA is part of comprehensive Public Financial Management (PFM) reforms in Poland. In July 2016, the CoM approved a six-point plan for budgetary reforms, based on proposals submitted by the MoF.8 As part of this plan, it was envisaged to eliminate the existing dual budget classification system. The objective is to introduce a new budget structure and improve data collection for budget and financial reporting, based on standardized and consistent public sector accounting rules applied through standardized classifications that are supported by upgrades of IT systems.
80. Since October 2016, the World Bank has had several technical discussions with officials from the MoF on the issue of introducing a revised SCOA and Budget Classifications. This work was supported by the Enhancement of Public Sector Accounting and Financial Reporting Program that came to an end in 2017. The aim was to broaden the scope of the CoA to address the reporting needs of the entire PFM system and to introduce one consistent and standardized classification structure across the government. The report comprehensively analyzed the four CoAs that are currently in use, discussed the conceptual issues related to the design of a SCoA, and made some recommendations for consideration in the design of a SCoA. These recommendations are presented in Box 5 below.9
81. The proposed SCoA reform has two phases, namely, the development of an integrated SCoA that will be followed by a second phase that provides IT reforms to support the implementation of the SCoA. This report lays out the scope of the first phase of the project and provides guidance on the workplan.
Box 5.Key Recommendations of the World Bank on CoA Reforms
The World Bank’s key recommendations were summarized as follows:
• “Development of a new integrated CoA which includes a range of segments to support different reporting requirements in government—the mission suggested that a new seven segment structure be considered; (These include Source of Funds, Administrative Organization, Sector/Function, Results/Program, Project, Geographic, Economic.)
• Development of a new economic segment that will better support budgetary, financial macro-fiscal, management and statistical reporting simultaneously—this requires benchmarking against a common international structure such as ESA2010 or GFS2014, identifying the gaps and issues, and then designing a more integrated coding segment;
• Presenting the budget on the basis of a fiscal balance which will in-turn directly feed into a focus on how the budget policies impact the balance sheet and net assets (even on a cash basis);
• Reporting entities should include a cash flow statement in annual financial statements. This is an area where Poland significantly departs from international reporting requirements;
• PFM reforms and supporting ICT systems should be developed with a much broader view to define requirements across the entire PFM system, not just from a central government perspective. The development of the budget classification (for budgetary reporting predominantly on a cash basis) and the four charts of accounts reflects a very narrow view of reporting and accounting requirements by specific units in the MoF. This in turn has led to fragmentation of the entire PFM system which should be addressed;
• Devolution of Authority and Responsibility has occurred without adequate policies and guidelines being issued— if MoF is to provide government assurance regarding the integrity of the PFM system, it should be able to provide direction and guidance to the reporting entities on a regular basis. Ideally, MoF should have the legal authority to issue clear policies, instructions and guidelines; and
• Fragmentation of ICT systems while an issue need not be viewed as an impediment to further reforms – the government should first focus on what is needed in terms of financial information and then address how ICT can provide the required solutions. In general ICT should not drive the reforms, instead it is ICT which should support the reforms.”
B. Key issues with the Current Chart of Accounts
82. There are a number of major issues with the existing budget classification and CoAs. These issues can broadly be summarized as organizational issues, lack of integration between the budget classifications and the CoAs, classification issues, and lack of standardization: (see also Box 3).
Organizational: There is no clearly defined administrative segment corresponding to accountability. In accordance with the Public Finance Act 2009, the budget is divided into 84 parts, corresponding to high-level organization units, sections of government administered by different ministers, types of spending or groups of public tasks, as well territories (as in case of the part of the voivods, which comprise spending relevant to different policy portfolios). Most accountability responsibilities are assigned to the holders of these budget parts. However, several budget parts could be organized administratively under one ministry, while at the same time important spending may be channeled through budgetary parts administered by holders other than relevant ministries. Therefore, responsibilities for government administration (such as policy formulation) and budgetary accountability responsibilities are not fully aligned.
Integration: The classification systems include four standardized CoAs and a budget classification that are not integrated at the state level. While the current CoA are consistent with the current financial statement requirements, the Polish financial statements differ in some aspects from the international standards. Budget information and financial statements are generated from the same basic accounting records and although it can be reconciled, the use of the financial statements remains limited and these statements are not consolidated at the state level.
Coverage: Various government entities are not included in the Trezor system10 and use their own idiosyncratic CoA based on various accounting policies. These include entities such as state legal entities, and various special funds.
Classifications: In the current budgetary classification, transactions are grouped in sections, chapters, and paragraphs. Sections and chapters were intended to reflect the functions of government, and paragraphs the economic classification. However, in its current format chapters mix organizations, programs, projects, activities and funds while the paragraphs mix elements of funding and cash flows. These classifications are also not fully consistent with the internationally comparable classification structures (such as the economic and functional classifications of GFSM 2014), and requires significant, difficult and sometimes impossible mappings with the international concepts. Furthermore, the current coding structure has limited digits, which does not allow for building a hierarchy in the structure in the budgetary classification.
Standardization: The various CoAs are standardized for particular types of entities, but are not standardized for all general government units. Concepts are not well defined, and in some instances, leave much discretion for interpretation. This hampers aggregations and consolidation across various government units with a reasonable level of confidence.
83. In addition, the current hierarchy of administrative units is unclear, with no clear link between low level reporting units and the higher level groups (such as ministerial portfolios). Identifying reporting units (called institutional units in statistics) and how these are related to each other should be resolved and made consistent throughout the fiscal cycle. Usually the structure of control between individual entities and the nature of their activities (whether they are market or nonmarket producers) determine how they are consolidated into groups (group reporting entities) and determine the sector classification in the statistical reporting. As explained in Box 2, the national definition of state budget and public finance sector, differs from the internationally agreed statistical definitions of budgetary central government unit/subsector and the public sector.11 While this approach, reflect domestic needs and specificity, it could create confusion with international users of the statistics. Such inconsistency between budgetary reports and the statistical reports could possibly make the reconciliation of data cumbersome and difficult.
84. While institutional coverage and sectorization issues fall outside the scope of this mission and the SCoA project, these issues will need to be resolved to inform the SCoA work. The identification of individual reporting units should be decided and aligned with assigned responsibility and accountability. For consolidation purposes, the design of the chart of accounts needs to identify items subject to consolidation. Attaining full consistency between the coverage of national reports and statistical reporting to Eurostat, OECD, and the IMF will increase confidence in the reliability of fiscal data.
C. Proposed Architecture of the Standardized Chart of Accounts
85. The World Bank proposal of a new seven segment integrated CoA forms a sound basis for the new architecture. The seven segments are:
86. The proposed segment structure will be standardized across agencies and will eliminate the current multiple use of the same segment for identifying various features of transactions. As indicated in Figure 4, the current structure of the budgetary classification often uses a segment for more than one purpose and therefore does not present data in a consistent manner for all transactions. In the current budgetary classification, the sections segment indicates for some transactions COFOG classification, while for others it refers to program classifications. Similarly, the chapter segment identify in some cases COFOG and in other cases the source of funding, while paragraphs are used to indicate source of funding, program, economic classification, or geographical classification. The current budgetary classification does not include a program segment—this program data are being compiled and provided as supplementary information to the budget. The proposed SCoA structure will integrate the current CoAs and budgetary classification and consistently identify all the proposed segments for each transaction and will include counterpart indicators. These proposed reforms will facilitate aggregation and consolidation of the general government sector.
Figure 4.Current Budgetary Classifications Compared to Proposed SCoA
87. While the proposed segments and levels within segments are largely in line with good practices, they could be further refined and improved before it is presented for agreement. Some areas include:
The proposed Source of Fund segment includes a level used to make a distinction between whether the source is a grant or a loan. This distinction is usually made in the economic segment and is therefore redundant here.
The program and project are presented as separate segments of the classification and could potentially be captured as sub-levels of the Administrative/Organizational segment.
The location segment does not provide for general headquarter based spending.
The levels of distinction in the economic classification (currently proposed to be five levels), could only be finalized after full analysis of user needs.
For statistical reporting needs, the distinction between transactions and other economic flows such as revaluations and volume changes in assets and liabilities is currently not provided for in the structure of reporting. This distinction could be built in as an indicator in the coding structure.
A distinction between current and capital spending could be built into the economic classification, as a sub-level not only of transfers but also for items such as compensation of employees and goods and services, which will allow the total of that aggregate to be derived from the system, but also the portion of it that needs to be capitalized in the cost of own-account capital formation.
To facilitate consolidation, the identification of counterparts to transactions should be built into the classification structure.
D. Issues for Consideration in the Development of the Standardized Chart of Accounts
88. A well-designed and well-structured SCoA provides the building blocks according to which all the potential and actual financial flows and positions of government are captured. As illustrated in Box 6, the SCoA should provide for the detailed information needed to aggregate, consolidate, or summarize into various reports that serve different analytical needs, including the budget classification. In an integrated system the SCoA would be supported by subsidiary systems that seamlessly interact with the SCoA to provide flexibility to capture even further levels of details to users as needed. For example, while a general ledger may record only the sum of transactions and stocks positions related to vehicles, a subsidiary system and/or asset management register could carry the details on each vehicle.
Box 6.The SCoA as Building Block to Align Budget Classifications and other Reports
Using the SCoA building blocks consistently in classifications for the budget, financial accounts, and statistical reports enables integration and reconciliation of budgetary, financial information, and statistical data.
This building block principle can be illustrated by subsidies:
The term subsidies is currently used in Poland budget classifications to describe all kinds of transfers that encompass items such as financial support to extra-budgetary units, capital injections into state-owned enterprises, transfers to support operations of loss-making entities, and transfers to support households exposed to some social risks. The implication of this is that
Financial Statements: For entities without large amounts of subsidies, the item will not be disclosed separately in the financial statements. Because it is not material it will be subsumed in an item for other operational expense. Entities with large amounts of specific types of subsidies, will have to disclose each type separately and come up with a term to describe it.
Statistical Reports: Each of the types of Poland subsidies the item would need to be reclassified into the appropriate economic category of expense: financial support to extra-budgetary units should be classified as grants to extra-budgetary units and will be flagged for consolidation with other central government units; capital injections into state owned enterprises should be classified as equity injections and will be flagged for consolidation of public sector units; transfers to support loss-making market entities should be classified as subsidies; and transfers to support households exposed to some social risks should be classified as social benefits.
In a SCoA, each of these types of transfers will be provided for in the correct subcategories of the economic classification structure and be defined appropriately. The category subsidies may be split into different types of subsidies to satisfy additional reporting needs as illustrated below:
Subsidies on goods
Public corporations ; etc.
Subsidies on services
Using the SCoA Budgetary classifications may be done at the aggregate level, i.e., subsidies, which will then be consistent with the total subsidies that will be reported in statistical reports. For financial statement purposes, the interest may be on making the split between subsidies to public corporations and they may report them separately, but the total subsidies will remain the same.
89. Creating an integrated SCoA is a balancing act between standardization and flexibility. Most time in developing a SCoA is spent finding the level of detail that should be included, the level of standardization that should be enforced, and the flexibility that should be allowed. Because different user needs require different levels of aggregations or presentational formats, the SCoA should be sufficiently detailed to serve the needs of as many users as possible, inform policy decisions appropriately, allow clear and direct linkages with sub-systems, and eliminate most of the needs to employ secondary collections/estimations of required data.
Flexibility to meet user needs
90. While broad agreement seems to exist on the need to develop consistent and integrated classifications of the budgetary accounts and the financial accounts, the full practical implications of the reform seem to be underestimated. The authorities were in general in agreement that a new economic, functional, accounting, and management classifications need to be implemented. However, in discussions on the implications of the reforms, various users of the CoA seem to want the change, only if their unique needs are served in the same way as in the past, and if the changes are not too extensive and costly.
91. The mission found that many of these “unique” requirements are indeed common across various departments. For example, they argued that they need to report on specific projects funded by the EU, or they need to serve specific management needs. It will be important to carefully determine user needs and plan change management activities to alleviate resistance to change during the project. Also, the level of standardization, and the level of flexibility that will be permitted should be carefully evaluated during the design phase.
Consistency and integration
92. As indicated in Figure 5, a well-built SCoA will serve to attain consistency and integration throughout the fiscal cycle. Various processes, procedures and reports will use and present fiscal information at different levels of details and various formats during the fiscal cycle. Using the same nomenclature and classifications in all these reports improve consistency, although they may be presented with different levels of detail.
Figure 5.Standardized Chart of Accounts Integration with the Fiscal Cycle
93. The way a SCoA could attain consistencies can be illustrated by the following example.12 An item such as compensation of employees may comprise several types of compensation for various categories of employment: for purposes of preparing forward estimates for the medium-term fiscal framework and budget, some of these elements may be influenced by different parameters and should therefore be grouped and estimated for each group. For purposes of presentation in the budget for approval, it may suffice to vote on the total of Compensation of Employees, while the financial plan and special purpose reports for departmental managers may require a subdivision according to the type of employment in order to manage employment costs effectively in the respective agencies. The SCoA allows for each of these levels of aggregations or disaggregation as needed.
94. Organizational issues around the reform, such as assigning roles and responsibilities and the timeline need to be firmed up. Attaining input and agreement on integrated SCoA reforms in Poland will impact a wide range of players. Clearly identifying roles and responsibilities, formally agreeing on the contributions each will make, and agreeing on timelines for deliverables would be a critical success factor in the development of the revised SCoA. These key players include:
Various departments in Ministry of Finance, among others, Accounting and Auditing, Expenditure Policy, Macroeconomic Policy, State Budget, Local Government, Financial Information, Tax and Legal;
Representative line ministries and other government agencies that will fall under the scope of the reform project;
Oversight bodies such as Parliament and the Supreme/Regional Audit Institution; and
External users such as the National Bank of Poland (NBP), Central Statistics Office of Poland, and regional/international organizations.
E. Pre-requisites, Assumptions, and Implications of the Standardized Chart of Accounts Reform Project
95. The development of the SCoA is part of a comprehensive program of PFM reforms approved by the CoM, and will be informed and influenced by other reforms (see Box 7). The SCoA reforms should therefore not be considered in isolation, as they will be dependent on various decisions in the other areas of reforms. In particular, decisions on accounting standards, principles and policies will influence the SCoA reforms, and IT reforms will be needed to support the SCoA reforms. At the same time, the reforms in the SCoA will have significant influence on other areas of PFM such as budget classification, presentation, and in-year reporting. High level agreement will be required for these decisions.
Box 7.Council of Ministers’ Decisions on Budgetary Classification Reforms
In June 2016, the Council of Ministers approved a Ministry of Finance proposal on the “Assumptions to Budget System Reform.” The document describes major directions of reforms and with respect to the improvement of the classifications of the budget the following is indicated as the objective:
“Elimination of existing classification dualism and introduction of new state budget structure and uniform performance-based classification (functional and program-based).
Efficient expenditure management requires elimination of classification dualism (traditional and performance-based), and the ensuing dualism of planning, execution and reporting. Uniform performance-based classification (functional and program-based) should be put in place to ensure efficient planning, prioritization and rationalization of expenditure. This classification should be integrated with the new structure of the state budget.
Classification should provide for systemization of expenditure in terms of:
function – corresponding to the competencies of respective ministries and state bodies which make up the basic structure of the state budget,
object – i.e. programs reflecting possibly uniform areas of public tasks and policies within the remit of ministries and state bodies,
subject – reflecting entities performing certain public tasks, and
economic – reflecting basic categories and groups of expenditure.
Public revenue will also require appropriate presentation in the new classification. Furthermore, it is necessary to analyze the best approach to tax expenditures i.e., reductions in government revenue due to preferential taxation.”
High level agreement
96. The development of an integrated SCoA requires high-level formal agreement to the SCoA reform process and sufficient broad political support. Establishing consistent classifications by implementing a SCoA could result in changes to the current presentations and high-level formats of various documents, not least of which being the budget classification.
97. Compilers and users alike should come to a common understanding of what the implications will be and be comfortable with that. The Council of Ministers decision on budgetary reform spells out the need for reform, but the implications of such unified and standardized reforms should be more carefully explored and communicated. A formal agreement between various Undersecretaries in MoF should underpin the reforms, and other senior role players and the CoM should regularly be informed on the progress with the developments.
98. Accounting standards, principles, and policies determine the development of a SCoA. While all entities are required to apply the Act on Accounting and some specific regulations on the templates of financial statements exist, accounting policies are not yet standardized in Poland. The MoF indicated that accounting is also guided by multiple legal acts regulating a variety of organizational and legal forms of entities in the public sector with many detailed regulations and some exclusions. These guidelines are applied in some cases in a cash basis of recording while in other cases it is applied to an accrual basis of recording. In addition, these allow for variances in application by various users. These multiplicities result in significant differences in the financial outcomes produced by various entities that complicates consolidation.
99. Decisions on adopting uniform accounting standards is pending the outcome of the work currently undertaken by Eurostat. This work aims at creating European Public Sector Accounting Standards (EPSAS) harmonized with International Public Sector Accounting Standards (IPSASs) for all EU members. Decisions on the SCoA will require convergence in the accounting guidelines and therefore possibly will have legal implications.
100. In lieu of the availability and adoption of EPSAS, a compromise decision on the adoption of uniform national accounting standards, regulations and policies will be required. IPSAS presents the most consistent solution, and while its adoption in the short term is unlikely, it does provide a primary reference point upon which the uniform national accounting standards and the SCoA can be developed. The existing accounting acts, principles and policies will need to be investigated and all discrepancies, contradictions and redundancies should be removed. Where existing standards allow a choice between various options, these would need to be removed, and any possibilities for misinterpretation should be eliminated by providing additional accounting guidelines.
Resources – staff
101. Developing a SCoA over a timeframe of two years is very ambitious and will require the allocation of sufficient resources to the project. Significant investment in terms of staff, time, and IT will be required. Resources with various levels of skills and knowledge, and various levels of decision-making powers should be committed and dedicated to the project.
102. Staff resources for the development of the SCoA should be organized by establishing various structures that could facilitate the work. At a minimum, they could be organized at three levels.13 The three main levels of structures are:
At a senior level, representatives from the various main role players should be formally organized to take the major strategic decisions related to the SCoA (for ease of reference the SCoA Committee). This group will mainly comprise senior staff from the budget, accounting, expenditure, tax, audit, and legal offices.
At a working level, a working group/secretariat with representatives from various role players involved in the development of the SCoA should be established. They should be dedicated to perform the various tasks related to the development work (for ease of reference the SCoA working group). This group will mainly comprise staff representing the budget, accounting and statistical compilers.14 The SCoA working group should execute the development work and serve as the conduit to collect and prepare information for the SCoA Committee and collect and inform the third level structure.
At a user level, either as compiler of accounts/data, or as a user of the data produced from the SCoA, a consultative group should be established (for ease of reference the SCoA users group). This group should comprise representatives from various type of units that will apply the SCoA either as a compiler or as a user.
Resources – IT
103. Usually, decisions on the conceptual design of the SCoA drive decisions on the reforms of IT solutions. However, in the case of Poland IT solutions within which a SCoA operates could potentially influence the design of a SCoA while in turn, a revised SCoA will require updates and reforms to existing IT systems. As indicated in paragraph 117, the various existing IT systems for accounting could easily be adapted to a new SCoA and will not put any restrictions on the design of the SCoA. The MoF has conducted a survey on the IT infrastructure currently in use. The MoF believes a revised SCoA will not impose significant additional demands on resources associated with existing IT systems. The implications for the IT systems that collate and consolidate the results will be scoped out during the first phase of the project, and will be considered in more detail in phase 2 of the EU project.
F. Workplan for Developing a Standardized Chart of Accounts
104. The World Bank Report elaborated on the core principles that should be considered in the development of a SCoA and described the various steps involved in redesigning the CoA. These principles and steps are based on a FAD publication15 and are therefore fully consistent with the views of the IMF. For purposes of project planning, the steps or tasks related to the project could be organized in various phases. These phases, with key decision points are presented in Table 1 and discussed in the remainder of this section.
|Phases||Key decision points||Responsibility||Indicative Timeline|
|Agree and sign Memorandum of Understanding between MoF Deputy Ministers||Nov 2017 – Jan 2018|
|Comply with pre-requisites||- Establish formally the organizational structure for project |
- Decide roles and responsibilities and agree on collaboration
- Agree accounting standard and principles that will drive the SCoA
- Allocate the relevant and dedicated resources to the project
|SCoA Committee||Nov 2017 – Mar 2018|
|Determine user needs||- Identify users/groups of users |
- Compare existing CoAs, and determine similarities and differences
- Survey existing analytical data needs and registers in use
- Determine additional user needs
|SCoA Working group in collaboration with SCoA users group||Apr 2018 – Oct 2018|
|Determine segments of SCoA and coding structure||- Agree segments to be implemented |
- Agree a coding structure to identify segments and main aggregates
|SCoA working group to propose and SCoA Committee to decide||Nov 2018|
|Seek approval of CoM on the conceptual design and implications of the SCoA design||Nov 2018|
|Develop detailed classifications of SCoA||- Determine line items based on consolidated user needs |
- Develop standardized definitions for line items
- Determine cutoff for standardized levels of detail
- Engage with users to review proposed classifications and definitions
|SCoA working group in collaboration with SCoA users group||Dec 2018 – Jun 2019|
|Test and revise||- Choose relevant agencies to engage in testing |
- Devise the modalities of testing
- Develop and test various report templates derived from SCoA
|SCoA working group in collaboration with relevant users||Jun 2019 – Dec 2019|
|Plan and agree a roll-out strategy||- Consider the impact on IT reforms |
- Consider the impact on relevant enabling instruments such as standards, laws and regulations
- Plan the development of accounting manuals and SCoA support systems
- Decide and agree timing and sequencing of roll¬ out
|SCoA working group to propose and SCoA Committee to decide||Jan 2020 – Mar 2020|
|Final approval of SCoA||- Approve and publish SCoA project outcome||SCoA Committee||Apr 2020|
Determining user needs
105. Determining user needs is a key step in developing a SCoA. Identifying all the user needs should take into consideration both the needs of the direct users, that is, those who use the SCoA to record economic events, and those of the indirect users, that is those who need information compiled from the result that were recorded in the SCoA. A comprehensive analysis will ensure that the SCoA serves the needs of all stakeholders. These stakeholders, who may have interest in the SCoA for various reasons, include:
Internal government users
Ministry of Finance (serving budgetary planning, oversight and accountability needs; accounting needs for the compilation of financial statements; and fiscal and macroeconomic data needs).16
Parliament, supreme/regional audit institutions and other government units.
Line ministries, legal entities, government managers of other government units.
Other domestic users (including Central Bank of Poland; Central Statistics Office of Poland; and Citizens, creditors and donors).
International and regional users (including Eurostat for reporting in accordance with excessive deficit procedure and the ESA 2010 transmission program; IMF, OECD, and World Bank for fiscal surveillance, economic analysis and adherence to the Special Data Dissemination Standard).
106. In determining user needs, the current reporting needs and the reporting obligations of the respective stakeholders should be examined carefully. Stakeholders consistently raised concerns that they may lose data that are currently available from their systems. By considering these current needs, and standardizing and organizing these in a consistent manner for all entities, data availability can be maintained. At the same time, the nomenclature and organization of such information may be accommodated in a different manner due to the standardization in the system. User needs should also encompass the introduction of common items that could fill gaps that may currently exist.
107. While a SCoA standardizes the structure and classifications at a certain level, it usually maintains flexibility at lower levels of the structure. Maintaining the flexibility to add unique sub-items is the only way to accommodate unique user need. However, meeting most of the common needs at higher levels will reduce the need for carrying additional directories. Also, meeting most of external user needs reduces the need to make use of secondary methods to fill these gaps.
Determine segments of Standardized Chart of Accounts and coding structure
108. User needs should ultimately determine the segments and broad categories of classifications that should be provided for in the SCoA. The tasks, goals and functions of various users and the information they need from the system will define the segments. While the World Bank’s proposed segments form a sound basis, the three primary classifications should be employed as separate segments, as they are essential for controlling, managing and reporting fiscal affairs. These are administrative, functional, and economic classification. Additional classifications such as program and project, fund and counterparts to transactions supplement these to create various types of reports. In each of these segments, various levels of details could be standardized while at some level of details reporting units could be allowed discretion to determine classifications.
109. A coding system and structure based on the agreed segments and group of classifications should be developed. Such a coding structure forms the linkage with the IT systems and provides the framework in which detailed classifications will be developed. This structure, with its associated nomenclature and definitions, forms the conceptual framework that will guide the development of detailed line items, and combination of items, that will feed into the various reports and presentational formats.
Develop detailed classifications of the Standardized Chart of Accounts
110. The agreement on the segments, coding, and major groups of classifications are then combined with the outcome of the user needs analysis to determine the detailed line items required for the SCoA. Three core principles should guide the development of the detailed line items:
Use current reporting lines as a basis, but find commonalities within additional itemized needs; match with international guidelines; decide common terminology to be used; and define the lines to eliminate misinterpretations;
Identify and eliminate duplications and redundancies, and determine whether any remaining data gaps exist;
Propose additional classifications to be introduced and needed;
111. Not all line items will be used by all entities. For the determination of the individual line items, given that a unified SCoA is the objective, some items will not be used by some entities. It is common for such items to be blocked for these entities to eliminate the chance that these are used by mistake. For example, local governments may have unique needs that will not be relevant for line ministries. For this reason, this phase of the project should also decide who will control the availability of various items. Often leaving that control to individual entities may result in misclassifications, but central control may be cumbersome.
112. As a final step in the development of the detailed SCoA, the working group should facilitate a review of the proposed SCoA by the users group. Such a review serves to identify line items that were omitted by oversight, but also helps to clarify nomenclature and definitions so that these become clear and not subject to misinterpretation.
Test and revise
113. The last step in developing a SCoA is to test the proposed SCoA. Testing the SCoA can be conducted in two manners. The most comprehensive testing can be done by introducing the SCoA in pilot entities, while maintaining the old classifications for control purposes. While this parallel run could be costly, it has the benefit that it most effectively and comprehensively test the new classifications in an environment where fear of losing existing information is eliminated. Once all tweaks are made and sufficient levels of satisfaction with the new SCoA is reached, the old system can be abolished. A second method to test the SCoA is to merely reclassify existing data into the new format and to test whether the classifications can produce all information needed for reporting templates.17 While less comprehensive and costly, this type of testing can be conducted as desk work with less demands on user entities, though it will only test the old user needs.
G. Future Work
114. Various tasks and reforms that are outside of the scope of the first phase of the SCoA project should be considered by the authorities. These reforms are brought about as part of the implications of the SCoA reforms. They broadly entail planning and preparing for the roll-out to all relevant agencies, reforming or updating IT systems and managing change during development and implementation of the SCoA.
Plan and agree a roll-out plan
115. A general strategy for the roll-out of the revised SCoA should be considered towards the end of the first phase of the project, with details to be worked out during the second phase. Implementing a new SCoA will require reforms of the relevant enabling instruments such as laws, standards, regulations, and policies to be considered, agreed and implemented. The reforms will impact on the processes, procedures, and presentation of information throughout the fiscal cycle. Therefore, based on these reforms comprehensive accounting, recording and reporting instructions and manuals should be developed. At the same time, the SCoA support systems, such as training and help desk functions should be established. Lastly, the timing and sequencing of implementing the SCoA in various groups of entities should be decided and be supported with the necessary allocation of resources for the reforms during implementation.
116. Implementing the SCoA will impact IT reforms, not only in MoF but more widely in all the compiler user groups. Currently entities can choose and contract their own decentralized accounting and reporting IT systems. At the level of the MoF, the Trezor system contains the budget, financial plans, and cash results of budget execution. This information is uploaded into the Trezor system from the result of individual accounting systems in all the State Budget units, except for the EU Funds. The SCoA will therefore require changes to both the Trezor System and the systems of all the entities that will be required to report using the SCoA. Decisions on whether the Trezor system will be able to facilitate the compilation of accrual based financial statements and consolidation are yet to be taken based on further IT assessment and reform initiatives.
117. An IT survey conducted by MoF in 2016 has indicated that existing systems will be able to cope with SCoA reforms. The survey conducted covered financial and accounting systems in 372 public sector entities and indicated that there are approximately 700 diverse IT systems in operation in these entities. These systems support various business processes and are provided by various vendors. Per the survey 93 percent of these systems have the feasibility to expand with new requirements. Of these, 85 percent requires changes by system providers, while 8 percent of them will be expanded by inhouse teams. Also, almost 20 percent of respondents plan to modify or replace their systems—most of them planned these changes for 2017. However, a full investigation of the implications of SCoA reforms on IT systems should be conducted. In terms of that, the envisaged project provides for a scoping mission to determine IT reform needs.
Change management and support
118. MoF should develop an appropriate change management strategy. Change management will be a critical success factor for implementing the envisaged SCoA changes, in light of the far-reaching implications that it has for reporting throughout the fiscal cycle. Key elements of such a change management strategy that should be considered are:
Prepare for change – including preparing an assessment of the readiness for change by various users;
Determine an appropriate communication strategy to introduce the change and the potential impact that it may have on users;
Determine information/training needs and strategies for various categories of users, including politicians, managers, technicians, and users of the output from the new SCoA system including civil society;
Obtain feedback – enforce success and introduce corrective actions during implementation.
119. MoF should establish appropriate support services for users of the system and establish procedures to update the SCoA to cater for new needs. Developing a new SCoA should be considered the starting point of reforms. Successful implementation will be determined by the level of support provided to implementing agencies. Developing supporting material, procedures to deal with queries, online and live helpdesk functions, and IT support should be established to ensure smooth implementation and functioning. Future changes to the SCoA should be considered very carefully to allow maintenance with the core principles agreed to in developing the structure of the SCoA, while at the same time reacting in a timely manner to emerging needs.
2.1 Investigate and decide the conceptual structure and process for development of the new SCoA
Review and decide on the SCoA segments to be included in the development of the SCoA based on user needs.
Consider and communicate the full implications of SCoA reforms to all stakeholders.
Decide and assign specific roles and responsibilities of various agencies in the development of the SCoA.
Formally agree on what reporting needs throughout the fiscal cycle will be considered in the design of the SCoA.
2.2 Carefully consider and communicate the pre-conditions, assumptions and implications of the SCoA project.
Communicate the practical implications of the SCoA reforms to high level decision makers.
Decide the accounting standards, principles and policies that will guide the SCoA reforms.
Allocate sufficient resources (staff and IT) in all relevant agencies to develop the SCoA during the envisaged two-year period.
2.3 Set up and staff a SCoA working group to:
Assess user needs in a comprehensive and inclusive manner.
Develop the framework and details of the SCoA based on these user needs.
Test and revise the draft SCoA as needed.
2.4 Consider and plan for future work related to the implementation the SCoA.
Plan and agree on a SCoA implementation strategy and its timelines.
Plan for IT reforms that will support the use of the SCoA.
Plan and execute change management activities that will support the implementation, maintenance, and use of the SCoA.
These are the State Budget, Local Government Finances, Paying Authority, Economy Financing and Budget Zone Financing departments.
The EC recommends limiting these decisions to “interventions by the government to change past policy orientations that is specified in sufficient detail, as well as adapted or at least credibly announced and has a direct incremental budgetary impact compared to the baseline.” See: European Commission, “Vade Mecum on the Stability and Growth Pact,” March 2017.
IMF Fiscal Affairs Department, “Poland: Developing a Medium-Term Budget Framework,” March 2017.
The paragraph grouping “current expenditure” excludes pension benefit payments.
IMF Fiscal Affairs Department, “Poland: Developing a Medium-Term Budget Framework,” March 2017.
Government controlled legal corporations operating as nonmarket enterprises may be the exception. EU laws may require a different accounting regime for them.
A CoA that are centrally decided and standardized for use by multiple agencies are often referred to as a standard chart of account (SCoA) or a unified chart of accounts. For ease of reference this report uses the term SCoA.
Ministry of Finance of Poland, “Assumptions to Budget System Reforms,” 2016.
World Bank, “Integration of the Charts of Accounts and Budget Classification,” May 2017 – for ease of reference referred to as the World Bank Report.
The Trezor system is an electronic system used by Treasury to collect reports from government agencies on their budget execution.
This problem is well illustrated by the National Road Fund’s exclusion from the state budget and public finance sector, while it is a nonmarket entity under the control of government that should be part of the central government and ultimately public sector. For statistical purposes, the fund is correctly included in the perimeter of central government, but it is excluded from state budgetary reporting and public finance sector data.
Using consistent and standardized classifications does not mean that all reports should be presented in the same manner. The SCoA should contain the necessary building blocks to allow customization of budget and other analytical presentations.
The second and third levels could be subdivided in smaller groups responsible for subsets of the work.
The staff resources from the Ministry should probably on average be available to dedicate 50–80 percent of their time to the project, while other members would probably need to dedicate between 20–40 percent of their time to the project.
These are based on the IMF Technical Notes and Manuals: “Chart of Accounts: A critical element of the Public Financial Management Framework,” Cooper and Pattanayak, 2011.
Using this option is limited due to the limitations that current classifications have.