FT Logo   Manual on Fiscal Transparency

IV.  Assurances of Integrity

149. It is essential for fiscal transparency that fiscal data reported by the government meet basic criteria that attest to their quality, and that there are mechanisms in place which provide assurances to the legislature and the public about data integrity. Principles and practices in this regard concern data quality standards, and public and independent scrutiny of fiscal data.

Data Quality Standards

4.1 Fiscal data should meet accepted data quality standards.

150. The Code includes good practices relating to: (1) budget data; (2) the accounting basis; and (3) assurances of data quality. All of these good practices are basic requirements of fiscal transparency.

Budget data

4.1.1 Budget data should reflect recent revenue and expenditure trends, underlying macroeconomic developments, and well-defined policy commitments.

151. Serious problems with budget execution—such as ad hoc short-term cash rationing, payment arrears, unappropriated expenditure, frequent supplementary budgets—can often be traced to poor budget preparation. A credible budget that functions as an effective tool of fiscal management requires at the outset that the revenue forecasts and expenditure estimates on which it is based are of high quality. It is a basic requirement of fiscal transparency that summary information on revenue forecasts and expenditure estimates should be provided in a background paper that is part of the budget documentation, and that detailed supporting information should be available for independent scrutiny.123

152. Realistic revenue forecasts are especially important since relatively small errors in forecasting this large aggregate can have a large impact on the fiscal balance given that expenditure is often difficult to adjust in response to revenue shortfalls. Revenue forecasts should be fully explained in terms of recent revenue trends, likely macroeconomic developments, and the estimated impact on revenue of any emerging consequences of existing tax policies and of any new tax changes. Countries will differ in their capacities to apply advanced revenue forecasting techniques. All, however, should indicate as precisely as possible the method used. Different approaches to revenue forecasting are outlined in Box 22.

153. On the expenditure side, reliability of estimates is more a function of rigorous costing and effective control mechanisms than of forecasting. Clearly, recent expenditure trends (which may reveal rising demand for certain public services) and likely macroeconomic developments (which will affect spending on interest payments and certain price or cyclically sensitive programs, for example, unemployment compensation) are important. However, the key requirement for reliable expenditure estimates is that they include all past and current spending obligations of the government. Persistent emergence of payment arrears is usually evidence that existing obligations are not fully covered in the budget, and therefore that there is a systemic data quality problem. Frequent and substantial use of supplementary budgets within year is likely to indicate that the original budget does not fully reflect the government's underlying obligations. Open budget procedures and thorough costing of budget policies provide a general assurance that supplementary budgets are unlikely to be needed.

Accounting basis

4.1.2 The annual budget and final accounts should indicate the accounting basis (e.g., cash or accrual) and standards used in the compilation and presentation of budget data.

154. Although there is no internationally accepted standard for government accounting or financial reporting, work by IFAC-PSC to develop such standards has progressed substantially, as indicated in Box 17. It is a basic requirement of fiscal transparency that reference should be made to the recognized or generally accepted accounting standards that are followed.124 Accounting policies should also be indicated.125 It should be clear where accountability lies within government for setting accounting standards and policies, and for monitoring and certifying compliance with standards. Any recent revisions in accounting methodology and practices should be disclosed, together with the reasons for the changes and an indication of their impact on fiscal aggregates (to facilitate comparability between years). Advance notice should be given of any significant planned changes in accounting policies or practices. Best practice is that mechanisms should be set up to provide for openness of the standard setting process for government accounting and financial reporting, and for its independence from government.127

Box 22. Revenue Forecasting

There are four main approaches to revenue forecasting.

Effective rate approach. Under this approach, the forecast for each tax is made by multiplying a forecast of the tax base by the corresponding effective tax rate. The effective tax rate is calculated by dividing the tax collected for the latest available period by the estimated tax base.126 Typically, the effective tax rate will differ from the statutory rate because of exemptions or incomplete taxpayer compliance. This approach can yield poor results when the tax base, tax rates, tax administration capacity, and taxpayer compliance are changing. For transparency, it is necessary to disclose the way in which the effective tax rate is calculated, the economic assumptions underlying the tax base forecast, and any adjustments that are made to reflect any of the aforementioned changes.

Elasticity approach. This approach establishes a stable empirical relationship between the growth in revenue for each tax and the growth in the corresponding tax base, which is specified as an elasticity. The increase in revenue is then forecast by multiplying the forecast increase in the tax base by the elasticity, and adding the estimated impact of changes in the tax structure and tax administration/compliance. For transparency, these components of the revenue forecast should be shown separately.

Model based approach. Some advanced economies use aggregate general-equilibrium models to produce revenue forecasts which take into account the interdependence of the tax system and the economy. Others use a sample of tax returns to build microsimulation models that describe the actual provisions of tax law, and use such models to produce micro-level forecasts that are then aggregated. These models can also be used in combination with the above two approaches. The effective rate approach or the elasticity approach can be used to produce a forecast on the basis of current policies, and microsimulation models can be used to produce estimates of the revenue impact of tax changes. Transparency requires that information on the models used, and various parameter values, are made available.

Trend and autocorrelation approach. In some cases, it is difficult to link revenue developments to underlying macroeconomic variables. This particularly applies to nontax revenue, which is linked to specific fees and charges, to profits of enterprises, or to property values. In such cases, past trends, supplemented by specific information related to each source of revenue, may be the only practical approach to forecasting. For transparency, the way in which the underlying trend has been determined should be specified, along with the relevant specific information that influences the forecast.


Assurances of data quality

4.1.3 Specific assurances should be provided as to the quality of fiscal data. In particular, it should be indicated whether data in fiscal reports are internally consistent and have been reconciled with relevant data from other sources.

155. The Code requires a public commitment to timely publication of fiscal information. The term "fiscal information" implicitly embodies a concept of quality. There is a presumption that any data published should be both reliable and relevant to fiscal analysis. However, it has become apparent that, in practice, more explicit standards need to be set to ensure that fiscal data are indeed of a high quality. For this reason, the Code has been modified to identify all data quality related aspects more clearly. Many aspects dealt with in earlier sections of the Code, such as classification, and coverage, timeliness, and periodicity, are aspects of data quality, but are sufficiently important in their own right to be dealt with separately. Fiscal data consistency and reconciliation is emphasized in this section of the Code.

Internal consistency

156. Crosschecks of internal consistency of fiscal data should be undertaken, and the effectiveness of these procedures reported. Fiscal reports, as indicated in Box 7, include the budget documentation, within-year budget reports, final accounts, financial reports, and GFS fiscal reports. It is essential that all of these reports meet high standards of reliability. It is a basic requirement of fiscal transparency that final accounts should be fully reconciled with budget appropriations, and that each should be reconciled with GFS reports. The latter provides assurance that all relevant accounts are covered by GFS reports. GFS reports should be compiled in parallel with budget reports, and should be actively used in the process of formulating and evaluating fiscal policy. Another basic requirement of fiscal transparency is that the change in the stock of debt (and financial assets) should be reconciled with the reported budget balance. Maintenance of a comprehensive government balance sheet is a systematic way of tracking changes in debt and assets, and can therefore provide a means of checking overall data reliability.

157. Assurance should also be provided as to the quality of fiscal data over time. For instance, where aggregate fiscal data are presented in the budget documentation for prior years (which is a good practice included in the Code), the status of the numbers (e.g., provisional or final) should be indicated. Any changes to the classification or presentation of items in the budget and fiscal reports from year to year should be disclosed, together with the reasons for the changes and their approximate fiscal consequences. Revisions to fiscal should follow a regular, established, and published schedule.

158. A further basic requirement of fiscal transparency is that a background paper should be included with the budget documentation which analyses the difference between budget forecasts of the main fiscal aggregates and the outturn for recent years. Best practice is that fiscal forecasts and outturns should be fully reconciled and all significant differences should be explained (preferably in the background paper mentioned above). In particular, differences between fiscal forecasts and outturns should be broken down into those due to macroeconomic factors, those that reflect the costs of existing policies, and those that reflect the costs of new policies.128 Where it is known that data are internally inconsistent, or that the reconciliation necessary to verify consistency has not been done, this should be clearly stated.

Reconciliation with other data

159. Reconciliation should be undertaken between fiscal data and related nonfiscal data, primarily monetary data but also balance of payments and national accounts data.129 It is a basic requirement of fiscal transparency that there should be rigorous reconciliation of fiscal and monetary data, and that where reconciliation processes are weak, this should be drawn to public attention (e.g., in audit reports) in a timely manner.130 Individual government ledger accounts should be fully reconciled with bank accounts. The overall balance measured as the difference between revenue and expenditure should be reconciled with domestic and external financing data as reported both by the government and by the central bank, the rest of the banking system, and other lenders. Financing data should also be reconciled with detailed information on changes in debt and financial assets. For all reports, any unexplained discrepancy between the government ledger accounts and bank accounts should be disclosed.

160. One way for countries to signal a commitment to improving the quality of fiscal data is to participate in the GDDS, and this is a basic requirement of fiscal transparency. A key purpose of the GDDS is to encourage member countries to improve data quality. The GDDS provides a framework for evaluating needs for data improvement and setting priorities in this respect. Participation requires, among other things, a commitment to using the GDDS as a framework for statistical development, that meta data are prepared describing current practices in the production and dissemination of official statistics, and that plans are announced for short and long-term improvements in these practices.

161. The general topic of data quality is also being dealt with systematically by the IMF through the development of a data quality assessment framework comprising a generic framework and a number of dataset-specific applications.131 A specific application for fiscal data is being developed that is consistent with the revised GFS Manual.132 The data quality assessment framework gives structure and provides a common language for the assessment of data quality. It is designed to be a flexible, comprehensive tool that can be used in a variety of country situations by experts and nonexperts alike. The framework aims to bring together best practices and internationally accepted concepts and definitions in statistics, including those of the UN Fundamental Principles of Official Statistics133 and the SDDS and GDDS.

162. A summary of the draft generic data quality assessment framework is presented in Box 23. The framework follows a cascading structure that flows from five main dimensions that have been identified as critical constituents of data quality. For each of these interrelated and somewhat overlapping dimensions, the framework identifies pointers, or observable features that can be used in assessing quality. These pointers to quality are broken down into elements (major identifiers of the quality dimensions), and further, into more detailed and concrete indicators (not shown in Box 23).

Box 23. Data Quality Framework—Main Dimensions

Prerequisites of quality

The legal and institutional environment is supportive of statistics; resources are commensurate with the needs of statistical programs; and quality is recognized as a cornerstone of statistical work.


Professionalism in statistical policies and practices is a guiding principle; statistical policies and practices are transparent; and are guided by ethical standards.

Methodological soundness

Concepts and definitions used are in accord with standard statistical frameworks; the scope of the statistics is in accord with internationally accepted standards; classification and sectorization systems are in accord with internationally accepted standards; and flows and stocks are valued and recorded to internationally accepted standards (basis for recording).

Accuracy and reliability

Source data available provide an adequate basis to compile statistics; statistical techniques employed conform with sound statistical procedures; there is regular assessment and validation of source data.


Statistics cover relevant information in the subject field; timeliness and periodicity follow internationally accepted dissemination standards; statistics are consistent over time, internally and with other major data systems; and data revisions follow a regular and predictable procedure.


Statistics are presented in a clear and accessible manner, forms of dissemination are adequate, and statistics are made available on an impartial basis; up-to-date and pertinent metadata are made available; and prompt and knowledgeable assistance to users is available.


163. The data quality assessment framework recognizes that the quality of an individual dataset, in this case, government finance statistics, is intrinsically bound with that of the institution producing it. This theme runs throughout the data quality assessment framework, but can be seen most clearly in the first two items in Box 23. Quality-of-the-institution issues with respect to the production of fiscal data are discussed in the next section.

123 See paragraphs 174-176 for a discussion of independent assessment of fiscal and macroeconomic forecasts.
124 For instance, IPSAS, GAAP, as in the United Kingdom and New Zealand, or Federal Financial Accounting Standards applied by the United States federal government (see http://www.fasab.gov/).
125 "Accounting policies are the specific principles, bases, conventions, rules and practices adopted ...in preparing and presenting financial statements;" see paragraph 7, IFAC (2000b). The accounting basis may differ between budget documents and financial reports, as it does for example in the United States. In France, accounting standards used in the preparation of the final accounts have recently been changed to reflect accrual principles in a number of areas, and these standards are clearly explained in the Final Accounts-see the ROSC for France, Fiscal Transparency, Box 1, at http://www.imf.org/external/np/rosc/fra/fiscal.htm. Some countries that have moved to accrual budgeting first went through a transitional period of reporting on an accrual basis while still budgeting on a cash basis.
126 Ideally, the tax base used should align closely with the category of tax collections but where such information is not available a broader measure, such as GDP, may have to be used.
127 For example, in the United States the Federal Accounting Standards Advisory Board is responsible for developing proposals to improve accounting and financial reporting in the Federal Government. In New Zealand, the Fiscal Responsibility Act requires the government to prepare and present all its fiscal reports in accordance with GAAP i.e. accrual accounting. GAAP is the responsibility of the New Zealand Accounting Standards Review Board, a body independent of the government that establishes accounting standards for the private and public sectors.
128 Australia provides a good example of routine accountability in these terms. See http://www.budget.gov.au/ finaloutcome/
129 In Albania, fiscal financing data are reconciled with financial sector claims on and liabilities to the government; and government debt and official flows are reconciled with the balance of payments. See "Toward a Framework for Assessing Data Quality," by Carol S. Carson, Annex IV, Sample C, at http://dsbb.imf.org/Applications/web/dqrs/dqrswork/.
130 In Pakistan, following a significant breakdown of the processes of accounts reconciliation, and the discovery of substantial fiscal data discrepancies, the government has begun to re-establish basic processes of accounting and reconciliation. The creation of an inter-agency Fiscal Monitoring Committee-and its strong support by the government-is an important step toward improving the quality of data used for monitoring budget performance. It has also resulted in a strengthening of internal reconciliation and control. See the ROSC for Pakistan, Fiscal Transparency Module, paragraph 26 and Box 1, at http://www.imf.org/external/np/ rosc/pak/fiscal.htm.
131 For detailed information concerning these framework, which was developed by the IMF's Statistics Department, see "Toward a Framework for Assessing Data Quality," by Carol S. Carson (2000), at http://dsbb.imf.org/Applications/web/dqrs/dqrswork/.
132 The fiscal data quality assessment framework is undergoing an intensive process of consultation with international experts and IMF staff, as well as field-testing.
133 See http://www.un.org/Depts/unsd/statcom/1994docs/e1994.htm. To further promote these principles, the UN Statistical Commission established a task force to develop a draft code of best practice. See United Nations Statistical Division, "Common Code of Statistical Practice in the United Nations Systems", Part I and Part II, April 1996 at http://www.un.org/Depts/unsd/demotss/tcnjun96/tony.htm.

Previous         Fiscal Transparency Manual Home          Next