From Governmental Accounting to National Accounting: Implications on the Portuguese Central Government Deficit

Based on the relevant differences between Governmental Accounting (GA/microeconomic perspective) and National Accounting (NA/macroeconomic perspective) this paper examines the main adjustments made in Portugal to the General Government Sector data required to convert Governmental Accounts into National Accounts. It also assesses the impact of those adjustments on the Central Government deficit, the largest share in the Portuguese public deficit. Following mostly a qualitative research methodology, the empirical study is based on interviews to officials preparing NA and on several documental sources. The purpose is to validate the major data adjustments from GA into NA regarding Central Government, while, in addition, assessing their impact using data from April 2008 Excessive Deficit Procedure notification, covering the 2004- -2007 period. The main findings indicate that differences concerning the accounting basis are the most relevant and that the subsequent adjustments have a considerable impact on the Portuguese Central Government deficit. This research points therefore to the need for more convergence between GA and NA, namely with respect to the transactions recognition criteria in order to use a common accounting basis, and for a complete and coherent reporting information system in GA.

Following mostly a qualitative research methodology, the empirical study is based on interviews to officials preparing NA and on several documental sources. The purpose is to validate the major data adjustments from GA into NA regarding Central Government, while, in addition, assessing their impact using data from April 2008 Excessive Deficit Procedure notification, covering the 2004--2007 period.
The main findings indicate that differences concerning the accounting basis are the most relevant and that the subsequent adjustments have a considerable impact on the Portuguese Central Government deficit. This research points therefore to the need for more convergence between GA and NA, namely with respect to the transactions recognition criteria in order to use a common accounting basis, and for a complete and coherent reporting information system in GA. Os principais resultados indicam que as diferenças de base contabilística são as mais relevantes e que os ajustamentos subsequentes têm um impacto significativo no défice da Administração Central portuguesa. Assim, esta investigação aponta para a necessidade de maior convergência entre os dois sistemas (CP e CN), nomeadamente quanto aos critérios de reconhecimento das transacções, por forma a ser adoptada uma base contabilística comum, e para um sistema informativo completo e coerente na CP. resumo résumé / abstract

Introduction
implementation of a Governmental Accounting reform, started in the early 1990s, is still in progress. The utmost step in this reform was the publication, in 1997, of the Chart of Accounts for Public Accounting (CAPA), which became the basic accounting framework for all governmental entities belonging to Central, Regional and Local Administration, except governmental business enterprises. Among others goals, CAPA aimed at «(...) obtaining essential elements in order to calculate the Public Administration values to National Accounts, very important to support the calculation and the assessment of the European Union Treaty convergence criteria» (Law-decree 232/97, Preamble, 7). This paper has the following goals: (1) to identify, from a conceptual point of view, the main divergences between GA and NA; (2) to examine the qualitative and quantitative relevance of these differences in the Portuguese context; (3) to analyse the main adjustments to be made when converting GGS data from GA into NA in Portugal; and (4) to assess the quantitative impact of those adjustments on the Central Government deficit reported by Portugal under the Excessive Deficit Procedure (EDP).
The paper is divided into seven sections. The first three sections include a literature review. Section 2 discusses the relevance of studying the relationships between GA and NA, focusing on the EU context, while Section 3 identifies the major differences between the two systems. Section 4 briefly explains the major features of the recent GA reforms in Portugal, in order to show whether the new system meets the ESA95 requirements. Section 5 presents the framework of GGS reporting under the EDP Notifications context. Section 6 explains methodological issues. The findings for Portugal are presented in Section 7, in the following order: relative importance of the main differences between GA and NA (subsection 7.1); major data adjustments when shifting from GA into NA (subsection 7.2); and the relative impact on the Central Government deficit reported (subsection 7.3). Finally, some conclusions and suggestions for further developments are presented in Section 8.
Council Regulation No. 2223/96 (and subsequent amendments) 1 forced EU member-States to adopt ESA95 in preparing their National Accounts, so that from April 1999 on the information sent to the European Statistical Office (EUROSTAT) would have to conform to the new system. Additionally (ESA95, §1.04), one of the specific purposes of this system is to support the control of the European monetary policy, namely national aggregates such as GDP, deficit and debt (EUROSTAT, 1996).
ESA95 is therefore the conceptual framework for EU member-States' National Accounts. In spite of the great diversity of political and social systems, underlying ESA95, it must reach not only objectives of analysis and evaluation of the economy of all member-States as a whole, but also to observe and control their fiscal and economic policies individually, in order to sustain the European Monetary Union (Sierra Molina et al., 2005).
Achieving these objectives implies, on the one hand, developing a real harmonisation of the (new) GA systems and, on the other hand, aligning those systems with ESA95 requirements, so that the macroeconomic aggregates are credible and comparable. The purpose is to get accurate values for the ratios established in article 104 of the European Union Treaty (1992) 2 and required Nevertheless, the problem of getting proper governmental sector data to National Accounts still remains. The data are obtained from Governmental Accounting, whose diversity and divergences from the macro accounting systems impairs the relevance, reliability and comparability of the aggregates that sustain the financial decisions of the EU member-States (Lüder, 2000 andSierra Molina et al., 2005). Montesinos and Vela (2000) explain that the main macroeconomic aggregates (e.g. deficit and debt) must have exactly the same meaning in governmental financial statements or in the National Accounts, otherwise usefulness and reliability of both accounting systems information may be significantly reduced. These authors also argue that governmental «(...) accounting information has a very important role to play in this process, as a useful tool for guaranteeing transparency and comparability among European countries and for economic, financial and management decision making» (Montesinos and Vela, 2000:129). Lüder (2000) adds that the course of GA reforms that have occurred in most European countries, with a tendency towards 'disharmonisation', has affected in a negative way the links to NA, increasing the inconsistency between the two systems and leading to procedures of 'creative accounting', derived from divergences between the scope of the reporting entities and the accounting basis. Montesinos and Vela (2000) also argues that certain practices of 'creative accounting' were adopted by several EU countries in order to offer a picture of convergence in the EMU, such as new methods for funding capital assets, delay in recognition of budgetary expenditures, and public debt decentralisation (public sector corporatisation).
The relationship between GA and NA has become relevant for several reasons (Cordes, 1996;Jones and Lüder, 1996;Lüder, 2000;Lande, 2000;Montesinos and Vela, 2000;Martí López, 2004;Sierra Molina et al., 2005;andBenito et al., 2005a, 2006): • The aggregates of NA related to the governmental sector are based on GA, so the convergence of these two systems is needed to assure reliability and accuracy of the output data that ultimately sustain the political decisions under de Stability and Growth Pact; • The figures on NA aggregates are the basis for the EU political and fiscal decisions; • The adoption of full accrual basis for the majority of transactions, according to the last revision of ESA95, compulsory for all EU members-States for preparing their National Accounts and from which the convergence criteria established in article 104 of the Maastricht Treaty (1992) are monitored 4 ; • The GA reforms in progress in several OECD countries, especially in the EU member-States, moving from cash-based to accrual-based accounting systems and considering the From Governmental Accounting to National Accounting: Implications on the Portuguese Central Government Deficit Maria Antónia Jesus Susana Jorge substantially and continuously and reach a level that comes close to the reference value; or alternatively, the excess over the reference value is only exceptional and temporary and the ratio remains close the reference value; whether the ratio of governmental debt over gross domestic product exceeds a reference value (currently 60%), unless the ratio is sufficiently diminishing and approaching the reference value at a satisfactory pace». controversial introduction of accrual-based budgets, given that the source of data from GA for NA comes essentially from budgetary systems 5 .
Though the studies about the relationship between GA and NA are rare, we have been analysing a few, including Cordes (1996), who focused his analysis on Germany, showing which improvements should be made in Governmental Accounting so that the data on the Governmental Sector meet ESA95 requirements. He emphasised, along with Jones and Lüder (1996), IFAC-PSC (2000) and Montesinos and Vela (2000), the following main differences between the two accounting systems: • Divergences related to the definition of the reporting entity under the concept of 'Governmental Sector', as in NA the 'Governmental Sector' only includes public sector entities producing non--market goods and services; • Differences related to the moment in which transactions are recognised, that occurs in a full accrual basis in the NA perspective (except for taxes recognition), while in the GA perspective both modified cash or modified accrual bases are still used as well 6 ; • Divergences related to the scope of the recorded transactions (the ESA95 requirement for recognising non-cash transactions, such as fixed assets depreciation, including infrastructures, not generally recorded in GA); • Differences related to measurement criterion of recognised transactions, which NA considers to be the market price while in GA the historic cost (acquisition or production cost) is preponderant.
• Jones and Lüder (1996), following Cordes (1996), extended his analysis to the divergences between NA (SNA93 and ESA95) and GA in two EU countries (Germany and United Kingdom), developing a comparative study and underlining the following main issues: • Contradictions between theory and practice the national aggregates computed under full accrual basis in NA and using modified cash basis in GA, and the implications on the determination of the ratios established by the EU Treaty; • Divergences raising from the reporting entity definition, based on the concepts of institutional unit and institutional sectors and subsectors according to ESA95, and taking into account that some public sector entities are not included in the GGS sector in NA; • Emphasis on the necessary adjustments to the figures provided by GA concerning the Governmental Sector due to different measurement criteria of assets and liabilities, reducing the reliability of the macroeconomic aggregates.
The differences between GA and NA might be identified from a conceptual point of view too, relating to different users' needs, which imply different objectives for the information provided by both systems. These differences imply others in the accounting principles adopted by each system, namely recognition and measurement criteria (AECA, 2001;Rodríguez Bolívar and Ortiz Rodríguez, 2002;Sierra Molina et al., 2005;Benito et al., 2006). Junho '10 / (24/46) 28 29

Differences between Governmental Accounting and National Accounting
We note that ESA95 general recognition criterion (full accruals) was later modified with respect to taxes and social contributions by the EU Parliament and Council Regulation (EC) No. 2516/2000, allowing member-States to use three different recognition methods: • Accrual basis -recognition when the taxes generating factor occurs; • Adjusted cash basis -recognition of taxes under cash basis sources, considering as much as possible, a time adjustment so that the amounts received can be attributed to periods when the economic activity generating the fiscal obligation occurs; • Cash basis -whenever it is not possible to apply none of the other methods.  Table 2 shows those that we believe to be the main issues and problems, among others, identified on that document as key differences between accounting and statistical basis of financial reporting as at June 30, 2004(IPSASB, 2005.
In short, our literature review allows us to identify the following major issues concerning the relationship between GA and NA: • Definition and scope of reporting entity under GA and NA; • Preparation and disclosure of consolidated financial statements; • Recognition of taxes and social contributions -tax credits, tax gap and moment of recording tax revenues 7 ; • Relationship between government and government business enterprises -privatisations, capital injections, government and government owned enterprises debt (notions of income and dividends).
Concerning the first issue, Lüder (2000) explains that while GA embraces almost all public sector entities (the exception being the Government Business Enterprises -GBEs), the GGS (S.13) under ESA95 includes only the institutional units that mainly provide non-market goods and services, intended for the general benefit of the community 8 .
On her turn, Lande (2000Lande ( , 2006 emphasises that NA collects micro data from several institutional sectors that present different accounting principles and criteria, making it necessary some adjustments in order to harmonise the moment in which the transactions are recorded and the measurement criteria that must be applied to those transactions. She also suggests the need for harmonisation of the conceptual framework of the accounting systems of all sectors of activity, including GA systems, in which adjustments are needed while shifting into NA, particular with Junho '10 / (24/46)

31
The scope of the reporting entity and the scope of the sector reporting Recognition of assets, except financial instruments Measurement of assets, liabilities and net assets (equity) • The boundaries of the reporting entities according to each accounting model • Reporting components of the public sector, namely of the GGS • Accounting for controlled entities • Definition of assets, liabilities, revenues, expenses and equity • Specific issues, such as costs associated with research and development, other intangible assets, extractive industries and defence weapons • Tax credits, tax gap and moment of recognising tax revenues • Public-private partnerships • Several different criteria for measurement of each item, such as impairment of non-financial assets, transaction costs, low interests and interest free loans, inventories, investments in associates and biological assets  (EUROSTAT, 2002) identifies the necessary conditions so that an institutional entity might be included in the GGS (S.13): the unit must be public (controlled by the General Government Sector) and non-market (sales cover less than 50% of the production costs). The Manual also defines how the concept of 'economically significant prices' (the 50% rule) must be applied to public institutional units in order to classify them into the GGS (EUROSTAT, 2002).
respect to the definition and scope of reporting entities and also concerning differences in classification, recognition dates and valuation methods.
More recently, Keuning and Tongeren (2004) emphasised the role of the adjustments, in a study of the relationship between GA and NA in Netherlands, in which they have highlighted the main steps that must be considered when taking data sources of Governmental Sector into NA: • Transformation of cash-based (GA) into accrual-based (NA) data -identifying the proper asset and transaction category; consolidating some internal flows; adjusting time of recognition of taxes, interest payments on central government debt, payments in advance, among others; • Transformation of accrual-based (GA) into accrual-based (NA) data -identifying the public entities market-oriented and also identifying the proper asset and transaction category and deconsolidate some internal flows.
GA reform in Portugal has followed the trend observed in other countries, namely in the EU, under the New Public Management context (Vela Bargues, 1996;Brusca Alijarde, 1997;Brusca Alijarde and Benito López, 2002;Brusca and Condor, 2002;Benito and Brusca, 2004;and Benito et al., 2005b) 9 : • Adoption of generally accepted accounting principles, namely introducing the accrual basis regime, with a progressive approach to business accounting; • Though the budgetary accounting system remains the most important, it has emerged as a subsystem of the Public Sector Management Information System; • Harmonisation of the accounting systems between different levels of government; • A more relevant role given to the accounting information for performance evaluation and management, as well as a tool for transparency and accountability; • Development of public sector accounting information, clearly oriented to decision making and to the evaluation of public sector efficiency and effectiveness; • Approaching GA and NA, so that adjustments, reclassifications and eliminations become easier and more reliable.
These changes in Portugal started at the beginning of the 1990s and many are still in progress. The main stages in the implementation of GA reforms in Portugal can be identified as shown in Table 3.  However, the landmark of the governmental accounting innovations in Portugal is CAPA, published in 1997. It was «(…) a fundamental step in the financial management and governmental accounting reform (…)» (Law-decree 232/97, Preamble, 1). This step was decisive to introduce accrual-based accounting in governmental entities, aiming at the «(…) determination of efficiency and effectiveness indicators related to the use of public money and, simultaneously, at the establishment of conditions required to prepare a Whole State Balance Sheet» (Caiado and Pinto, 2002: 40).
According to Carvalho et al. (1999), GA should continue with traditional objectives of legal control associated to budgetary accounting but, at the same time, it must include recent purposes related to assessing economy, efficiency and effectiveness in decision making, implying the introduction of financial and cost accounting. Caiado and Pinto (2002) also explain that until the publication of CAPA, in 1997, GA was understood as a set of rules for the budget approval, accomplishment and control, in order to judge the corresponding responsibilities. But this traditional budget-oriented governmental accounting, concerned with an overall purpose of legal accountability of the sources and uses of financial resources, with a single-entry booking method, neither has allowed evaluation of the patrimonial and financial situation of reporting entities nor has provided useful information for management.
The main goal of CAPA was «(...) to create conditions for the integration of different accounting perspectives -budgetary, patrimonial and cost accounting -in a modern governmental accounting system, which is a fundamental instrument to support governmental entities management and evaluation» (Law-decree 232/97, Preamble, 6).   , 7), covering the budgetary and financial/patrimonial perspectives and the National Accounting point of view as well.
The following CAPA major characteristics can be underlined (Caiado and Pinto, 2002): • Though continuing considering issues related to budgetary accomplishment, the accounting system became more oriented to patrimonial, economic and financial issues.
• Both financial and budgetary transactions are recorded under a double-entry bookkeeping method; • Financial and budgetary accounting, although using distinctive accounts, are integrated in one single accounting system, that should also integrate cost accounting; • Full accrual basis for financial transactions and modified cash basis for budgetary transactions.
Therefore, the CAPA integrated system uses simultaneously two accounting bases in a clear divergence with the NA system -ESA95 which is full accrual-based (except for transactions regarding taxes and social contributions) and demands for Balance Sheets showing the sectors net worth and respective changes. Divergences may also be found in other areas, for example concerning the assets measurement criteria -historical cost (CAPA) versus market price (ESA95), or the identification of the reporting entity.
Moreover, using the annual activities reports of the Portuguese Public Administration Accounting Standardisation Commission (CNCAP, 2005(CNCAP, , 2006(CNCAP, , 2007 it can be seen that CAPA has been implemented almost exclusively in the autonomous services (with administrative and financial autonomy), while the administratively autonomous services (only with administrative autonomy) still apply a modified cash-based budgetary accounting system. This CAPA implementation problem emphasises the divergences to NA (essentially all accrual-based), considering the still relevant weight of administratively autonomous services in the Portuguese Central Government Sector, which makes the cash-based accounting still predominant.
Since 2001, Portugal has adopted a new State Budget Framework Law, as well as new rules for preparation and presentation of the annual budget (according to projects and programs) and for accountability of certain public entities under the control of several institutional bodies -e.g. The Court of Accounts (Supreme Auditing Body) and The Portuguese Republic Legislative Assembly (Parliament).
The reforms implemented in the second and third stage (c.f. Table 3) are still in progress, but it is important to underline that they have not included the accrual approach for the budgetary system, which is considered an important limitation to this reform process at the present stage. The legislative changes in this field so far have merely followed the international trends, introducing new practices for budgetary transparency, according to OECD (2001) recommendations, and new measures to accomplish with the requirements of Stability and Growth Programmes which all European member-States under the EMU are obliged to.
The main legal establishments for the Excessive Deficit Procedure (EDP) Notifications concern the application of the Protocol on the EDP, as well as both ESA95 and ESA95 Manual on Government Deficit and Debt (EUROSTAT, 2002), with the subsequent updates 10 .
Nowadays, it is very relevant the so-called 'Institutional Cooperation Agreement' (INE et al., 2006)  accomplishment of EU legislation about GGS statistics and the methodological harmonisation concerning data collection and compilation, as well as the update of sources and methods used in the preparation of National Accounts regarding the GGS.    As our study is focused on Central Government, Table 7 identifies the main data sources used in the non-financial accounts, half-finalised and final data (planned and estimated data come only from DGO budgetary estimates), in April Notification, regarding Central Government accounting data 11 .
11 In NA, non-financial accounts refer to flows from transactions between institutional units related to expenses and revenues. The balance of these transactions relating to GGS is the governmental deficit/surplus, meaning the net borrowing/lending of Sector S.13. On the other hand, financial accounts deal with financial transactions (borrowing and lending) taking place between institutional units and show how the deficit is financed (or the surplus applied) by financial assets and liabilities. Otherwise, financial transactions are related to acquisition or amortisation of financial assets and liabilities (e.g. borrowing and lending; acquisition and sale of shares and other financial investments). Interest paid/received is however a flow included in non-financial accounts.
EDP Notifications periodical reports are prepared not only from different data sources but also using different accounting bases: for the half-finalised accounts all data are cash-based, except for the entities of the Ministry of Health, while the final accounts are cash-based for the State and accrual-based for SFAs.  Thus far we have followed mostly a qualitative methodology to describe, analyse and compare accounting practices, focalising on a particular context and pursuing a systematic and integrated approach (Miles and Huberman, 1994;Ryan et al., 2002).
One may also argue that a case study method has actually been used, as our approach allowed us to describe the accounting systems and examine the techniques and procedures in their practical setting, being a fieldwork applied to a specific country instead of a particular organisation (Ryan et al., 2002;Yin, 2003).
Qualitative studies sometimes use qualitative and quantitative data together (Miles and Huberman, 1994). In the former, several sources and research techniques have been used, such as analysis of documents, as well as semi-structured interviews, following the research lines designed by Yin (2003).
While the literature and the documental review were helpful in framing the problem and identifying the divergences between GA and NA, the interviews were designed for, say, validation purposes of the major differences across the two systems for the Portuguese case, as well as understanding their conceptual and quantitative relevance. The interviews were made to officials responsible in Portugal for the preparation of NA and EDP Notifications, from both INE and DGO. The semi-structured interviews were taped and occurred between December 2007 and March 2008.
Regarding the BdP, also responsible for the National Accounts, namely regarding the computation of the public debt, it was only possible to get written answers to a few questions related to data compilation and information about the quantitative relevance of the differences mentioned above.
The interviews were developed into two steps. Firstly, several preliminary and informal conversations occurred in order to get information about the major problems regarding the differences between GA and NA in Portugal. In a second step, we designed a guidance document comprising the most important issues to be treated in six main interviews to both INE and DGO officials, with an average length of 90 minutes each: (1) the relationship between GA and NA systems in Portugal; (2) the areas where differences between GA and NA are more significant and the reasons for specific accounting treatments; and (3) the determination of conceptual and quantitative relevance of differences previously identified.
The analysis of official documents, such as the Inventory of Sources and Methods (INE, 2007), allowed us, in a third stage, to identify the most important data adjustments (from GA into NA), namely concerning Central Government.
Finally, to assess the impact of those adjustments on Central Government deficit reported under EDP context, we used quantitative data, collected from

Main differences between GA and NA in Portugal
The interviews made at INE and DGO, and the written answers from the Bank of Portugal, allowed us to identify the major differences between GA and NA and their relevance from each institution's point of view. Table 8 summarizes this information, and the interviews outcomes confirm the main differences identified on the literature review, which can be grouped in three categories: (1) the scope of the GGS; (2) the accounting bases and (3) the capital injections in State-owned corporations.

Methodological issues
The first one is the most relevant but, similarly to category 3, it does not imply standardized adjustments procedures, only occasional reclassifications in order to meet the requirements of ESA95 Manual on Government Deficit and Debt. This issue concerns the definition of the boundaries of the public sector entities in the institutional Sector S.13 (GGS) according to the Manual rules. These rules establish that a governmental entity must be included into the GGS only if it is a non-market institutional unity (which production is sold at economically significant prices) 13 . In Portugal INE is the institution responsible for applying these criteria, defining which governmental entities are to be included in the GGS.

Data adjustments from Governmental Accounting into National Accounts
The differences in the accounting bases involve data adjustments from GA into NA, made consistently in order to transform cash-based data into accrual-based data. The Inventory on Sources and Methods (INE, 2007)   2516/2000. For primary expenditures, the adjustments convert cash data (cash + commitments) into accrual data, based on information collected from a questionnaire sent to all public sector entities involved in cash-based accounts. Interest adjustments are made by the Portuguese Government Debt Agency (IGCP) that analyses loan by loan and security by security at the individual level in order to apply ESA95 rules. Table 9 shows several adjustments for these situations, concerning accounting bases differences.
The Inventory of Sources and Methods (INE, 2007) also distinguishes adjustments related to occasional reclassifications of some transactions, which require complementary information that is not available in the accounting records (i.e. General State Account or Financial Statements). Table 10 identifies those transactions, the data sources and the corresponding adjustments.

Impact of the accounting differences on the Central Government Deficit
The quantitative impact of the accounting differences on the Government deficit is assessed analysing data of   In general terms, Figure 1 presents the comparison between the total amount of all adjustments and the Portuguese Central Government deficit (from GA, subsector State, before adjustments), showing that adjustments are greater to final and half-finalised accounts. Because these data incorporate more adjustments, they are more accurate, whereas estimated data integrate only a few initial adjustments. Therefore, the final Central Government deficit (   There are several divergences between Governmental Accounting and National Accounting from a conceptual point of view, namely concerning the criteria for transactions recognition (e.g. cash basis versus accrual basis) and measurement (historical cost versus market price). Differences can also be found in the scope of the reporting entity and in other relevant specific issues, such as those related to the preparation of consolidated financial statements and to the relationship between Government and Government Business Enterprises. Most of these differences have implications on the determination of macroeconomic aggregates.
Regarding Governmental Accounting, the reform process in Portugal has generally followed the practice of other European countries, moving from cash to accrual basis. Subsequently, it might be said that the convergence with macro accounting has been improved. But major differences still remain in place, namely in measurement criteria (historical costs versus market price), but also regarding the existence in GA of two different accounting bases -accrual basis for financial accounting and modified cash basis for budgetary accounting.
In turn, from the interviews to Portuguese officials involved in the preparation of NA related to GGS and the analysis of several documental sources, one may conclude that the main differences between GA and NA in Portugal are related: a) to the scope of the GGS; b) accounting basis (cash versus accruals), c) reclassification of certain transactions, such as capital injections in State-owned corporations. Based on the Inventory of Sources and Methods (INE, 2007), we have also identified the major adjustments related to the conversion of GA data into NA.
The quantitative impact of the reported differences on the Central Government deficit was based on the Portuguese EDP Reporting for Central Government (April 2008 Notification, TABLE 2A). The analysis of the impact in the 2004-2007 period has shown that the accounting basis differences became materially more relevant. It follows then the need for more convergence between the two systems -GA and NA -namely regarding the recognition criteria in order to use a common accounting basis.
This study is an attempt to quantify the impact on Central Government deficit arising from the transition from GA into NA. Further developments are of course required, as well as a more detailed analysis of the accounting basis differences. It would also be interesting to evaluate the impact of the accounting differences between the two systems on the Portuguese Central Government debt by analysing