Public Sector Consolidated Statements—an Assessment

Governments have experimented with, or produced, whole of government consolidated financial statements on an accrual accounting basis for more than three decades. The US Treasury prepared prototype consolidated financial statements in 1976 (Bowsher, 1988, p. 42) and have published annual reports in that format since 1995 (and audited statements since 1997).¹ Similarly, Canada has been producing its ‘public accounts’ in the form of consolidated statements since 1995 (though initially these had a limited scope and only reflected the partial application of accrual accounting).²

In Australia, a government-appointed ‘commission of audit’ in the state of New South Wales produced consolidated reports for the year ended 1987, and these were formally published by that state after 1989 and subject to full audits from 1993. By the end of the 1990s, the profession had issued an accounting standard AAS 31, Financial Reporting by Governments (June 1998), while the Commonwealth and all state governments were publishing consolidated statements.

New Zealand has published whole of government consolidated statements since 1993, apparently on the basis of procedures developed by N.Z. Treasury. Some European countries have reportedly prepared variants of consolidated statements since the late 1980s.³ In the U.K., HM Treasury proposed to produce unaudited whole of government accounts for 2003–04, and a full set of audited accounts in 2005–06 (HM Treasury, 1998). This followed publication by the U.K. government of a Code for Fiscal Stability (1998) that was subsequently given a statutory basis in the 1998 Finance Act. The code stated that the government shall ‘produce accounts for the whole public sector and that these should be produced on a consolidated basis where reasonably practicable’ (HM Treasury, 1998, p. 8). However, implementation has been deferred.⁴

Despite this experience, the literature on public sector consolidation accounting is sparse. Two decades ago it was suggested that ‘in the few cases where governments actually prepare comprehensive financial statements, there are no generally accepted accounting standards for them to apply’ (Egol, 1988, p. 181). Certainly, until recently there has only been limited discussion of the theoretical rationale for the production of whole of government consolidated reports. Most contributions have come from professional bodies, notably from the Canadian Institute (CICA, 1980), the Australian Society of CPAs (ASCPA, 1993), an Australian Accounting Research Foundation (AARF) monograph (Micallef et al., 1994), and a series of studies produced by the International Federation of Accountants (IFAC, 1991. 1993, 1996). To this brief list can be added the U.K.'s HM Treasury discussion paper on whole of government accounts (1998), and the U.S. Federal Financial Accounting Standards Advisory Board (FASAB) discussion paper on issues associated with the preparation of whole of government consolidated statements (FASAB, 2003).

As noted above, an accounting standard on public sector consolidated statements was issued in Australia in 1998. The U.S.A.'s FASAB issued a standard (SFFAS 24, 2003) which nominated what previously issued standards were to be applied in the consolidated financial report of the U.S. government. This preceded the issue of Statement of Federal Financial Accounting Concepts No. 4, Intended Audience and Qualitative Characteristics for the Consolidated Financial Report of the United States Government (FASAB, 2004). Mention might also be made of the contributions of the U.S.'s Governmental Accounting Standards Board (GASB), whose Statement 14, The Financial Reporting Entity (1991) which, while not dealing directly with public sector consolidated statements, established guidelines for the identification of the scope of financial reports for state governments. The GASB prescribed alternative forms of multi-column presentations of information about ‘primary government’ and ‘financially accountable component units’ in a ‘fund accounting’ context. These modes of reporting are similar to multi-column consolidated statements as are commonly used in the private sector (in that case, to present financial information about parent and group entities). The above-mentioned IFAC discussion papers can be regarded as precursors to attempts to standardize practice, at least for those countries that adopt accrual accounting in financial reports for the public sector. IFAC has issued two standards, IPSAS 6, Consolidated and Separate Financial Statements (2000), which is intended to apply to both the public and private sectors, and IPSAS 22, Disclosure of Financial Information About the General Government Sector (2006).

These various accounting standards and discussion papers are in several important respects inconsistent. This article reviews and assesses the general case for the presentation of consolidated statements encompassing whole of government, general government (or other sets of governments or government agencies). The main focus is on the identification of the objectives of this form of reporting, and hence the delineation of the reporting entity (or area of consolidation) for public sector consolidated statements, and the scope of accompanying consolidations of sub-sectors of the public sector. The analysis proceeds by considering claims about the potential users of aggregated financial data about the financial position and performance of governments, and the judgments likely to be made on the basis of that information and whether those judgments are routine, or whether the need to make those judgments would only arise under certain conditions or in exceptional circumstances. It is concluded that a number of routine judgments could be based on financial information presented in the form of consolidated statements but that these would not necessarily be consolidated statements encompassing the whole of government (or, to be more specific, statements encompassing the whole of a specific tier of government). Prior studies or standards have prescribed or recommended that the test of control be applied to determine the scope of public sector consolidated statements (e.g., IFAC, 2005; and standards IPSAS 6, 2000; AASB 127, 2004; AASB 1049, 2007). This article proposes alternative criteria for determining the scope of consolidated statements.

While accounting standards in some jurisdictions have required the preparation of consolidated statements for individual agencies,⁵ or local governments,⁶ the focus here is on the use and design of consolidated statements for nations or states. The term ‘public sector consolidated statements’ is used to relate to aggregative reports dealing with any set of public sector agencies within a particular tier of government (national or state). The term ‘whole of government accounts’ is used to apply to consolidated statements that purport to deal with all agencies within either a single tier of government (national or state) or, in some circumstances, several tiers of government.

WHY PRODUCE WHOLE OF GOVERNMENT CONSOLIDATED STATEMENTS?

While there is widespread support for the proposition that governments should produce whole of government financial statements on an accrual basis, there remains disagreement in the advocacy literature as to the target readership of those reports, and what judgments they are intended to inform—matters which would shape the scope of those reports and the technical practices to be undertaken in their compilation.

IFAC (2000) summarized much of the debate by suggesting that there were four approaches to the identification of a public sector reporting entity: depending on whether the focus of attention was to be on ‘the authorized allocation of funds’, the ‘legal entity’, ‘political accountability’ relationships, or ‘control’ (p. 16). It neither prescribed nor recommended particular treatments; nor did it present a detailed analysis of the specific objectives these differing concepts of whole of government reporting might serve—beyond acknowledging that adoption of one rather than an another approach could produce different results. Rather, IFAC offered a generalized statement, namely that ‘the overriding objective of financial reporting is to communicate reliable information which is relevant to the decision making and accountability needs of users’ (1996, p. 2). However, there may be inconsistencies between the objectives of providing public or private sector information to assist decision making, and providing information relevant to accountability (Walker, 2003). Without some analysis of how potential users might seek to use financial information for different purposes, it is not possible to maintain the claim that both objectives could be accommodated in a single set of financial reports.

Curiously, most published commentaries have promoted the preparation of a single set of consolidated statements encompassing the whole of government. Yet those governments which have been preparing public sector consolidated statements have often chosen to accompany whole of government reports with suites of consolidated reports which encompass separate sub-sets of government agencies, disaggregated in terms of institutional sectors, for example as defined in Standard Economic Sector Classifications of Australia (ABS, 1998) for the purposes of Government Finance Statistics. If there is a case for whole of government reports, there may also be a case for alternative or supplementary presentations—yet the relative merits of the alternatives have only been minimally debated.

The structure adopted here proceeds from an initial consideration of both the target audience (existing or potential) of public sector financial reports, and the way that information may be used or be useful. In other words, it considers who are the users of those reports, and what judgments they may make on the basis of the information contained in public sector consolidated statements. But before doing so, some prior literature is reviewed to indicate how there does not appear to be a unified view on either who are the potential users of these reports, or on what are the potential uses of that information.

WHO ARE THE POTENTIAL USERS?

There is a distinct contrast between the claims made in studies published by the accounting profession, and those produced by independent academic researchers.

A pioneering U.S. study by Anthony (1978) presented a list of the predominant classes of potential users of public sector accounting reports (while acknowledging that such a list reflects the judgments of the person who compiles it (p. 41). Later authors have commonly appealed to the authority of prior studies—occasionally adding another group, or rearranging lists of users into new classifications. Jones et al. (1985) provided a convenient review of a series of studies of this genre, and then explained that they had developed their own list of primary users by examining selected studies and accounting literature.⁷ Other authors have relied on literature-generated lists of users to sustain normative proposals about the design or content of financial reports. For example, Sutcliffe et al. (1991) referred to the authority of AARF's SAC 2 to identify the main categories of users; they then asserted that particular categories of users ‘will be concerned’ with certain issues, or ‘will require’ particular kinds of information (pp. 29–31)—without offering any evidence or analysis in support of these assertions. That enabled the authors to claim that the need of users would be satisfied by ‘general purpose financial reports’—a circular argument, since a ‘general purpose financial report’ was defined in SAC 2 (1990) as a financial report that was intended to meet the information needs of those same major classes of external users.

IFAC (1991) was able to assert that it could list the ‘principal users of government financial reports’ on the basis of ‘a review of accounting literature and authoritative pronouncements on financial reporting by governments’ (p. 7). Similarly, the Australian profession's AARF claimed that there was a consensus about the identity of users of public sector reports (in general) (AARF, 1991; Micallef et al., 1994). This assertion was based on a compilation of the opinions of eight writers— notwithstanding that most of the opinions collated were those of ‘preparers’ rather than users.⁸ Accordingly, the claim that there was a ‘consensus of opinion’ only reflected the choice of the works cited, and the way opinions had been presented. Indeed, the eight works surveyed by AARF did not include any of the sixteen studies on the same topic previously cited by Jones et al. (1985).

Other projects aimed at obtaining evidence of the use of public sector financial reports by members of the public have failed to produce compelling evidence about readership (see Rutherford, 1992). In one study, researchers received a zero response rate to a simple questionnaire left in copies of a local government authority's annual report at three library sites, for a total period of twelve months (Butterworth et al., 1989). Similar observations about minimal usage of these reports were provided by Crook (1993). Jones (1992) offered the startling observation that ‘the publication of financial statements is not in the public interest because the public has no interest’.

Some limited evidence about the identity of users has come from surveys administered to different classes of respondents. Indeed, AARF itself once disclosed that it was undertaking a survey of users of public sector financial reporting by government departments. AARF later was compelled to reveal that responses to this survey were ‘disappointing’ (AARF, 1991). When those results were reported, it appeared that all responses (only thirty-seven) came from preparers and auditors—with not a single one from an external user (Sutcliffe et al., 1991, p. 28).

As noted above, Anthony (1978) explained that the intention was to only refer to potential users, a stance echoed by other academic contributions (e.g., Mayston, 1992), and later by the U.S.A.'s FASAB whose SFFAC 4 (2004) referred to the ‘intended audience’ of consolidated financial reports, and identified citizens and citizen intermediaries as ‘primary audiences’.

It seems fair to summarize commentaries or research findings regarding the identity of users of public sector financial reports as follows:

Table 1 summarizes some statements about the supposed users of public sector financial reports. The summary is based on statements made by a range of government agencies or professional bodies (or studies commissioned by those bodies) and identifies nine groups of nominated users.

In context, the list of potential users is partly derived from illustrations of the membership of supposed user groups. For example, a joint U.S.–Canadian study by Dye and Bowsher (1986, p. 10) referred to three groupings: legislators and government, citizens and corporations, and the media and analysts. IFAC (1991) identified five principal classes of users: legislative and other governing bodies; the public; investors and creditors; other governments, international agencies and other resource providers; and economic and financial analysts. FASAB (2004) suggested that users of financial information about the federal government can be classified in four major groups: citizens, Congress, executives, and program managers. There seems little point in quarrelling with these assertions about ‘classes’.

But classification is a purposive activity. It appears that the purpose of these contributions is to support claims that certain forms of financial information would meet the needs of main classes of users. That is a matter of evidence. The authors were essentially making assertions without providing that evidence. And, of course, before public sector consolidated statements were actually being prepared, evidence could not be obtained from observations of practice.

Some major differences in the claims made by different contributors may not be immediately evident in Table 1. For example, a commonly expressed view is that public sector consolidated reports should be aimed at external users. In a study commissioned by the GASB, M. Ives, then Director of Research of the U.S.A.'s GASB, noted (in a preface to Jones et al., 1985) that members of the executive branch of government were not considered ‘primary users’ (p. iii). Similar views were expressed in an Australian Accounting Research Foundation (AARF) study (Micallef et al., 1994). A 1993 IFAC report noted that ‘internal reports required by management or special reports’ were designed to meet the particular needs of specific users, and the information likely to be contained therein ‘is too diverse to be dealt with effectively’ (p. 1). Yet FASAB (2004) and HM Treasury (1998) saw public sector consolidated statements as informing internal users, such as government planners, program managers, or managers of administrative units.

Another major difference is evident from the way that several contributors have claimed that individual governments—all tiers of government—should be regarded as separate reporting entities. Indeed, this view was reflected in Australian accounting standards AAS 27 and AAS 31. Yet the HM Treasury, alone amongst this set of contributors, saw whole of government accounts as revealing information about a government's national fiscal policies—with the implication that public sector consolidated statements should be prepared to encompass secondary (or, for some countries, tertiary) levels of government, local or state government entities, and bodies that ‘exercise functions of a public nature’ or that are ‘entirely or substantially funded from public money’ (HM Treasury, 1998, p. 4).

WHAT ARE THE POTENTIAL USES?

Notably several standard-setting bodies have engaged in advocacy of their preparation and publication of public sector consolidated statements before examining in any detail the judgments likely to be made on the basis of data obtainable from those reports (or from other sources). This is despite prior acknowledgement that such an exercise is desirable. Indeed, some contributors to the advocacy literature seem to have oscillated, suggesting that a short-cut to implementation was warranted. For example, Anthony (1978) observed that the identification of users and their needs was an important element in the standard-setting process—though later he noted that exercise ‘does not help in determining the principles that should govern the preparation of financial statements’ (Anthony, 1989, p. 106). And that the onus was on others to demonstrate that the needs of users of the financial statements of public sector and other non-business entities were such that private sector accounting was not appropriate—but that he was ‘not aware of any such demonstration’ (p. 26).

Similarly, Sutcliffe (1985), in a study commissioned by AARF, outlined a series of issues ‘to be resolved’ before the profession started issuing accounting standards in the public sector. The issues identified: ‘Does the identification of users as “recipients of services”, “providers of resources” and “other parties” provide a suitable basis for consideration of financial information disclosures to be made by public sector entities?’ (p. 26). However, when some survey-based research undertaken by AARF was inconclusive (so that those issues had definitely not been ‘resolved’), AARF then claimed that empirical research was unnecessary as there was a ‘consensus’ about the information needs of users (AARF, 1991, p. 3).

Overall, the literature analysing the uses which are (or might be) made of different forms of financial reporting—particularly public sector consolidated statements— has suffered from defects similar to those in the literature concerned with the identification of users. Many published commentaries on whole of government consolidated reporting have simply presented the case for that form of reporting. They recognized few difficulties, beyond that of initially implementing accrual accounting throughout the public sector, and in resolving a series of technical issues (e.g., regarding the valuation of heritage or infrastructure assets). Further, while concluding that consolidated statements would constitute general purpose reports of relevance to a variety of users, they failed to demonstrate that argument—by indicating that items of information were of common relevance to a range of users in relation to a range of regularly confronted judgments.

More significantly, many published commentaries have failed to recognize that existing arrangements for financial reporting already produce a wide range of information relevant to many of the judgments faced by their list of supposed users. One exception was the study by Micallef et al. (1994), which acknowledged the existence of budget papers and other reports but then dismissed them as inadequate because they did not meet the AARF definition of general purpose financial reports.⁹

At best, such claims about the information needs of certain classes of users can be restated as hypotheses about what types of information are relevant to those users. That restatement can then be amplified as a series of hypotheses, for example, that one might expect to find a specified type of information used by individuals when making certain types of judgments in certain situations, if it was available; or that in certain situations, individuals who have to make judgments of certain kinds would search for certain types of information if it was not routinely made available to them—and so forth.

Further, a more detailed analysis of the uses made of particular kinds of information would recognize that not every category of user may directly access detailed financial reports, but rely either on summary documents or on intermediaries to interpret complex information for them.

Moreover, the existing literature has given little or no recognition to the possibility that the way in which information is used may be contingent on external events. In other words, certain types of information may be used or sought in some situations (such as when a government is recording persistent cash deficits), but not in others.

Arguably a fundamental weakness of several profession-sponsored studies is that claims about information needs were couched in generalities, and were presented without any detailed analysis of the types of judgments faced by individual users, and how different types of information would be relevant to those judgments. In some instances, simple assertions about how presumed users may employ accounting data failed to explore the manner in which the data might actually be used. For example, the following assertion sounds appealing, even plausible:

But if one dissects the statement, it is exposed as virtually meaningless. First, financial information on its own is inadequate to enable a reader to ‘assess a government's performance’ in relation to service provision. Any such assessment would also require information about the services provided, and to whom, and with what effect.¹⁰If ‘financial information’ is to be used in this evaluative process, it may well be used as the numerator or denominator of a series of performance indicators about issues relating to economy (e.g., trends in material costs per unit of service) or efficiency (e.g., cost per customer, cost per unit of service provided). It is unlikely that financial information about ‘government’ will be of relevance to assessments of effectiveness (e.g., the impact of a preventative health campaign on the well-being of a community; or the extent to which an educational campaign improved literacy or numeracy amongst schoolchildren).

In summary, the existing literature on the kinds of uses which might be made of public sector financial information:

Without careful consideration of the prospective uses to which public sector consolidated financial statements might be put, financial reporting practices are unlikely to furnish information that is optimal, in the sense that it can be regularly used to inform judgments by a wide range of stakeholders.

ALTERNATIVE APPROACH TO THE DESIGN OF CONSOLIDATED STATEMENTS

An effort was made to consider the way in which actors or observers of the public sector may refer to aggregative statements about financial aspects of the activities or circumstances of governments. The analysis follows the following four steps:

The analysis was undertaken by referring to institutional arrangements in Australia, and recognizing the existence of a three-tiered structure of government (Commonwealth, state and local government), even though the focus is on reporting by national or state governments.

A total of eighteen separate judgments were identified. Observations about these judgments are arrayed in Table 2. Some of the items were identified by simple observation of institutional arrangements: for example, the existence of a price regulatory body with jurisdiction over public trading enterprises (such as the New South Wales Independent Regulatory and Pricing Tribunal) indicates that judgments are being made about the pricing of government services. Other items were identified by reference to the text of political debate or analysis: if a politician or commentator discussed the capacity of a government to service debt, that implies that judgments are being made about such an issue.

The identification of potential judgments was based on a similar process. The categorization of judgments as routine, conditional or exceptional was based in part on observation and partly on speculation. Likewise, the specification of ‘relevant data’ is based in part on the text of political debate, lobbying activities, and press comment. The listing of relevant data remains partly speculative, and is certainly partial. The listing of ‘relevant reports’ refers to materials published by one or more Australian governments or government agencies—with the exception of references to various forms of public-sector consolidated statements (shown in bold type). The listing of different forms of consolidated statements reflects an argument: that those reports would convey relevant information to the nominated users in relation to specified types of judgments.

Some explanations are provided regarding the eighteen judgments summarized in Table 2:

1. Results and sustainability of a government's financial management practices: While media commentary tends to dwell on past budget results and projected budget outcomes, this emphasis on deficit or surplus represents an inappropriate ‘search for a single measure of fiscal impact or some other relevant aspect of budgetary and financial policy’ (Buiter, 1990, p. 2). In Australia, the customary presentation of these results has been in the form of budget papers and periodic (monthly or quarterly) reports on budget results. At the state level in Australia, the reporting entity for these documents has variously been the fund used to handle recurrent payments, the consolidated fund, the consolidated fund together with certain trust funds, or (more broadly) a set of entities described as the ‘budget’ or ‘general government sector’.

Claims about the relative performance of states of New South Wales and Victoria became a political issue during a state election in N.S.W. in 1991 (see Walker, 1995c, p. 113), focusing attention on this diversity in the identification of the reporting entity for budget reporting. For example, the state of Victoria's budget presentations during the late 1980s encompassed entities providing public housing and public transport, and operating area hospitals—while other states did not include these in their consolidated fund results. One state managed its state finances through three major funds (a Consolidated Revenue Fund, a General Loan and Capital Works Fund, and a Trust Fund) and its annual budgets related to only the first of these (see Independent Commission (W.A.), 1993, Vol. I, pp. 19–33).

Participants agreed at a May 1991 Premiers' Conference to move towards standardization of state budget presentations in line with the Australian Bureau of Statistics (ABS) Government Finance Statistics basis.¹² However, the states and territories took a considerable time to secure uniformity (see, e.g., Walker, 1993). Indeed, some states appear to have manipulated reported budgetary data (and budget results) by selectively excluding agencies from the scope of budgetary reporting¹³—or perhaps some state treasury departments believed that alternative presentations were more useful for their purposes. Whatever the reasons, the ABS has undertaken material adjustments to the data published by the states for its publication Government Financial Estimates Australia (Cat. 5501.01). Arguably, ABS data are more consistent and reliable than the data provided in budget papers or other documents produced by individual states.

While advocates of public sector consolidated statements have tended to focus on reports encompassing whole of government, the high level of political and media attention devoted to budgets and budget results suggests that data from a consolidation of general government agencies would potentially be regarded as highly relevant. Where budget papers are prepared on an accrual basis, they may include estimated financial results for the current year; however, the consolidated statements would be compiled at a later date (and would be audited).

Note that Table 2 lists parliamentarians as potential users of this information, but this must be qualified. In practice, parliamentarians, as a class, are not actively involved in making decisions about resource allocation. Rather, budget strategies are usually crafted by a small group of political leaders and their advisers, and bureaucrats from central agencies. Most members of government have little say in the matter (see, e.g., Nicholls, 1991a). Indeed, in Australian governments, many ministers are ‘budget-takers’ not ‘budget-makers’. As for members of the Opposition, they generally play no role at all in these processes—beyond offering comments in the course of parliamentary debate, or seeking to obtain political advantage from information gleaned after the event when published budgets are considered by Estimates Committees.

2. Whether a government has been successful in implementing its overall financial policies: A wide range of parties may endeavour to make assessments of the outcome of specific policies, though reports on financial activities (disaggregated by categories of expenditure, or programs) will only be one of several sources of information relevant to those assessments. Cash-based reports may well be the dominant source of relevant data for assessing the likely impact of fiscal policies (e.g., promoting employment through capital works programmes) or assessing the outcome of plans to raise funds from specific forms of taxation. Other sources of information include special-purpose publications (see, e.g., N.S.W. Government, 1989), or become available from performance audits undertaken by public-sector auditors, or reviews undertaken by parliamentary committees.

If a government raises funds from user charges (such as for water or electricity services provided by PTEs) the outcome of such activities may be reflected in part in budget papers showing dividends and other transfers of funds from those agencies. (The consideration of pricing policies is referred to below—see judgment #6.) However, regard may also need to be had to the extent to which infrastructure held by those PTEs has been maintained or upgraded (see judgment #5).

3. Capacity of a government to continue to deliver existing levels of services to the community (or to enhance these services) at existing levels of taxes and charges: Judgments about these matters may be made by parliamentarians, their advisers, and by public sector managers. Interest groups frequently make pre-budget submissions, or later refer to budget documents, in order to promote or criticize government policies regarding expenditure in areas such as welfare, education, health and public housing (see, e.g., NCOSS, 2003) or on road funding or regional development (see, e.g., Australian Local Government Association, 2006). In some instances, interest groups may publish proposals in order to influence the policies of competing political parties in the lead up to elections (e.g., NCOSS, 2007).

A comprehensive assessment of the capacity of governments to maintain or enhance services may require not only information about the overall financial position of a government, but also about the cash flows required to meet those obligations (such as pensions or health care costs) not recorded as liabilities—a point made strongly in the U.S. environment by Walker (2004). Arguably, decision makers should also have regard to information about the physical state of infrastructure required to deliver those services (see judgment #4), and about the projected costs of introducing alternative equipment or processes. All of these financial and non-financial factors would need to be assessed against projected needs arising from demographic changes. Assessments in relation to specific services may be based on information provided by the agencies delivering those services—though (if the information is available) such assessments may be made in the context of the government's likely cash flows in the medium to long term.

4. Capacity of a government to continue to provide infrastructure which will maintain or enhance existing levels of service to the community: Within government, judgments about the scale of allocations of funds to capital works may be determined by the premier and treasurer, who establish the broad parameters for a budget. While members of parliament may be invited to nominate projects for capital works, decisions about the allocation of funds to specific projects are generally made by the budget committee on the advice of treasury departments, or special advisory committees set up for that purpose (see, e.g., Nicholls, 1991b; N.S.W. Treasury, June 2008; N.S.W. Government, 2008). Government budget papers generally include a separate volume dealing with capital works.

On occasions, interest groups advocate greater expenditure on capital works (see, e.g., AusRAP, 2007; ACOSS, 2006; NCOSS 2008). Proposals to cut existing levels of service, or to vary prices so as to create incentives for reduced consumption, may also promote discussion within the community on infrastructure. Parliamentary committees and officers with an oversight role have promoted the idea of developing state-wide plans for future investment in infrastructure (see S.A. PAC, 1987, 1988; N.S.W. PAC, July 1993, N.S.W. Audtor-General, 2005) and such proposals have been implemented through the development of ‘state plans’ (see, e.g., N.S.W. Government, 2006; N.S.W. Treasury, June 2008) or policies on ‘total asset management’ (N.S.W. Treasury, May 2008). When such exercises are undertaken experience suggests that political imperatives may prevail, as published plans highlight capital projects that are expected to have electoral appeal, often with little or no attention being devoted to expenditure to remediate or upgrade legacy infrastructure.

5. Performance of a government in maintaining the infrastructure required to deliver services to the community: Present reporting arrangements by governments do not include the presentation of assessments of the physical condition of infrastructure assets (for a review of reporting options, see Walker et al., 2000). Reports dealing with the management of infrastructure assets have been provided by state Public Accounts Committees (see, e.g., S.A. PAC, 1988; N.S.W. PAC, July 1993) or public servants (e.g., Woods, 1991). Some studies have suggested that a lack of coordination of the plans of individual agencies could lead to peaks in demand for capital to fund asset replacement (e.g., S.A. PAC, 1987). The Commonwealth Auditor-General has advocated that ‘annual budget papers should highlight infrastructure spending undertaken on an annual basis so as to enable the public to determine the extent to which future generations are being asked to pay for the recurrent operations of today's generations’ (ANAO, 1993, p. xiii). In response, the Commonwealth Treasury noted that there was no accepted definition of what constituted infrastructure (as opposed to capital works; the Department of Finance noted that ‘the substantial majority of public sector capital outlays occur at the State/local level’). The ANAO acknowledged difficulties but observed that there is presently ‘little accountability for the physical condition or effectiveness of use of infrastructure which can lead to doubt whether infrastructure expenditure is maintained at appropriated levels in the generational context’ (p. xiv). Politicians or media commentators have also referred to problems arising from backlogs of maintenance (e.g., N.S.W. Commission of Audit, 1988; N.S.W. Government, 1990; N.S.W. Auditor-General, 2001).

6. Manner in which a government is pricing services: Some governments have articulated pricing policies for services provided by NFPTEs (e.g., price increases will be limited to increases in the Consumer Price Index (see e.g., N.S.W. Treasury, 1991, 2001) Other governments have established agencies to review or recommend pricing practices for their own agencies (e.g., U.K. Oftel, Ofgas; N.S.W. Independent Pricing and Regulatory Tribunal).

While PTEs generally are treated as being outside the ‘budget sector’, government budget papers commonly include commentaries on the performance of those enterprises. Further, special-purpose reports describing the performance of PTEs and trends in prices over time have been issued by governments either individually (see e.g., N.S.W. Treasury, 1992, 2005) or collectively (see SCONPMGTE, 1993; SCRGSP, 2005). Reports on the performance of PTEs have also been published by the Industry Commission (1990, 1992), the Economic Planning Advisory Committee (EPAC, 1992) and the Productivity Commission (2005).

The presentation of budget papers that encompass both government departments and public-sector enterprises appears to have been first adopted by a territory government (see Northern Territory Budget Paper No. 2, 1992–93). Others have followed suit.

7. Extent to which a government is subsidizing the cost of services that are sold to the community: The extent of subsidization has been addressed in a range of articles or special purpose reports dealing with specific services or industries (see, e.g., Whiteman, 1988; Bureau of Transport & Communication Economics, 1989; Brew, 1991). During the 1990s a series of reports from parliamentary committees examined the cost of community service obligations (e.g., Victorian Economic and Budget Review Committee, 1991; JCPA, 1992; N.S.W. PAC, January 1993), while the Steering Committee on National Performance Monitoring of Government Trading Enterprises produced a report on the cost of government business of providing community services (Steering Committee, 1994).

Some Australian governments adopted the practice of identifying CSOs and then ‘funding’ them directly from budgetary allocations (see N.S.W. 1993–94 Budget Paper No. 2). It has been claimed that the aim of this practice is to ensure that managers of government agencies face ‘clear objectives’ (N.S.W. Treasury, 1993). However, the process of identifying and then ‘costing’ CSOs may involve arbitrary assessments (Walker, April 1992). Moreover, an activity (such as providing rail freight for grain crops) might be regarded by one side of politics as core business, and by the other side as a CSO. Possibly classification as a CSO means that a particular service is subject to ongoing scrutiny of how they are defined, funded and delivered (see e.g., Queensland Treasury, 1999, National Competition Council, 1999; Western Australia Treasury 2000).

All general government agencies are (by definition) primarily reliant on budgetary allocations to provide services—but they do not necessarily sell those services to individuals within the community. On the other hand, some government trading enterprises (e.g., those engaged in public transport) may also be primarily reliant on subsidies (variously described as operating grants, CSOs, or pensioner concessions) but the extent of these subsidies may not always be apparent from the annual reports of those agencies

8. How a government has spent taxpayer's funds (and any borrowings): Governments have traditionally produced annual budget papers, and periodic financial reports (monthly or quarterly) which summarize the financial results of activities, on a cash basis (or modified cash basis).¹⁴

These statements form the basis of most media comment on government financial activities. Indeed, even though different Australian state governments have adopted widely different assumptions about the scope of the reporting entity when reporting their budget results, most media commentators appear to take the figures at face value, and even offer observations on the relative merits of budget strategies, on the basis of data compiled in quite inconsistent ways.

The cash-based reports of past financial results which are presented in annual budget papers are unaudited. Subsequently governments publish the public accounts or treasurers' annual financial statements which have been examined by the auditor-general. However, under present arrangements these public accounts of many states relate to the results of the consolidated fund and trust funds rather than to the whole of the budget sector or general government. Moreover, when published, these reports contain old news and attract little attention unless the auditor-general discloses departures from parliamentary appropriations (see judgment #10, below).

In 1993–94 the N.S.W. Public Accounts for the first time covered the Budget Sector and were prepared on an accrual basis. Subsequently other states and the Commonwealth have followed suit.

9. Whether a government is incurring obligations which will impose burdens on future generations: Arguments about inter-generational equity have mainly concentrated on the build-up of liabilities—a phenomenon which has not been reflected in cash-based reports on budget results. Such issues have been addressed by Auditors-General (see, e.g., ANAO, 1993). Parliamentary committees (see, e.g., N.S.W. PAC, 1984), or special reports, often commissioned by incoming governments (see, e.g., N.S.W. Commission of Audit, 1988; Victorian Commission of Audit, 1993; Report of the Independent Commission, W.A., 1993).

Commonly, the significance of figures for aggregate liabilities (or net liabilities—net of financial assets) is considered by reference to benchmarks, such as Gross National Product or Gross State Product (see, e.g., Nicholls, 1991a).

10. Whether there has been compliance with parliamentary appropriations: The Westminster system of government requires that allocation of resources to agencies be subject to parliamentary oversight. IFAC (1991) claimed that a common objective of presenting information was to enable assessments to be undertaken of ‘whether resources were obtained and used in accordance with the legally adopted budget’ and ‘whether resources were obtained and utilized in accordance with legal and contractual requirements, including financial limits established by appropriate legislative authorities’.

Some jurisdictions provide some latitude by automatically providing supply for a limited period, or by providing advances to the treasurer (to meet the cost of ongoing programs prior to the budget bills going before parliament, and thereafter for certain contingencies), or through arrangements whereby certain expenditure can be undertaken without prior parliamentary appropriation, provided the treasurer brings the matter before parliament at a later date. If breaches of these requirements occur (either by the Department of Treasury or by individual agencies) the matter may be the subject of reports by the auditor-general to parliament. Correspondingly, an auditor-general may report on departures from parliamentary limitations on borrowings by individual agencies, either directly or in substance through use of complex financial transactions.

11. Feasibility of budget policies: Prior to the preparation of an annual budget, the formulation of budget strategies and policies is customarily undertaken by an inner circle of executive government (such as an executive review committee, with the advice of public servants from central agencies (Departments of Treasury, Finance, Cabinet, and Premier or Prime Minister). The existence of these committees is commonly reported in the media.

12. Attractiveness of investing in debt securities issued by governments: Potential investors in government securities will have regard to the yields available from that investment, relative to that available from alternative investment opportunities. Credit rating agencies assign ratings to debt issued by governments or individual government agencies.

The credit rating of governments may become a political issue and the subject of ongoing commentary. Before the availability of accrual accounting data, rating agencies attempted to estimate the value of major liability items (see, e.g., S&P Australian Ratings, December 1992).

In the Australian context, a key factor affecting the revenues of the states and territories is the manner in which the Commonwealth allocates funds (via general purposes or tied grants). As one credit rating agency observed, ‘the strong financial support provided by the Commonwealth government through fiscal transfers to all states’ is a key factor in establishing the rating of individual states (Moody's Investors Service, Credit Opinion, 11 October 2007).

Because of the reliance of the states on these allocations, in practice a state's credit rating cannot be higher than the Commonwealth's. Allocations of funds to the states and territories are undertaken on the advice of the Commonwealth Grants Commission, and are subject to limited negotiation at annual conferences of the heads of Australian governments (formerly termed Premiers' Conference, later described as the Council of Australian Governments, and including a representative of local government).

Another key factor is the extent to which states can borrow further funds. Borrowing arrangements have been regulated by the Loan Council, through the establishment of global limits for borrowings by individual states after the Commonwealth has determined the overall allocation. (For a history, see Senate Select Committee, 1993.) However, these arrangements have been changed, with a new system of Loan Council allocations (see Loan Council, 1993).

The ratings agency Standard & Poor's has claimed that it ‘rates about 190 sub-sovereign governments in the developed world outside the United States and only 16% are rated AAA’ (Submission to Senate Select Committee on State Government Financial Management, 19 March 2008). At the time of making this submission most states and the Australian Capital Territory were rated AAA.

13. Attractiveness of maintaining investment in debt securities issued by government: The market price of government securities establishes yields to redemption. In most cases, a downward revision of credit ratings might signify that a ratings agency considered that there might be some delay in redemption, rather than doubts about ultimate repayment.

In the Australian environment, the limitations on state borrowings established by the Loan Council have imposed a discipline that has enhanced the attractiveness of government paper. (However, see discussion below in relation to judgment #15.)

14. Risk of delays in payment to creditors: As noted above, there may be little risk that governments may default in payment of their debt; there may be greater risk of delays in repayment. For those dealing with many agencies, that risk may be remote. However, some agencies (such as some New South Wales area health boards) have periodically experienced cash shortages leading to complaints from suppliers about delays in payment (though arguably these difficulties arise from financial management practices by bodies responsible for delivery of politically sensitive services, rather than the ultimate capacity to pay by government-funded agencies).

In some jurisdictions efforts have been made to impose fiscal disciplines on government agencies by the inclusion in annual reporting rules of provisions for reports to be provided on performance in paying accounts (see, e.g., N.S.W. Annual Reports (Departments) Regulation 2005, Schedule 1, ‘Report of Operations’).

15. Oversight of levels of borrowings by a regional (state) government: The Loan Council arrangements adopted after 1993 provide for quarterly reporting to the market, supposedly to ensure transparency. However, the proliferation of forms of off-balance sheet financing designed to avoid the restrictions imposed by the Loan Council, created challenges. Initially the Loan Council proposed fuller disclosure of exposures to financial arrangements involving private sector financing of projects, but this was resisted and by 1995 this commitment to disclose contingent exposures was abandoned (for a review of this history, see Walker and Con Walker, 2000).

These financing arrangements included BOOT schemes (arrangements whereby a private sector consortium builds, owns and operates a tollway or other physical infrastructure, and then transfers it back to government at the end of the defined term). These may start out as unperformed executory contracts but at least some of these arrangements may evolve into firm financial commitments once infrastructure has been constructed, and are in many respects similar to finance leases. The failure of accounting standards to require disclosure of the scale of what may be, in substance, liabilities, remains a limitation on the usefulness of public sector consolidated statements as a source of information about the scale of overall public sector indebtedness.

16. Extent to which the public sector is drawing on national savings: It has been claimed that deficits in the public sector are at the expense of private savings and investment (Egol, 1988, p. 179) or involve a draw-down of public savings. Such concerns may shape long-term policy by national rather than state governments, and policy judgments about these matters are likely to be made with regard to statistical data concerning economy-wide savings and investing activity—not from accounting reports. However, recent commentary suggests that sustained large budget deficits may have a larger impact on long-term economic growth due to their effect on market expectations and a related loss of confidence ‘both at home and abroad’ (Rubin et. al., 2004).

17. Financial circumstances of (Australian state) governments (vis à vis other Australian state governments): Commentaries on the financial circumstances of state governments are published in the media, while periodic comparative surveys of state finances have been published by bodies such as the Institute of Public Affairs and (for more than a decade) by the Evatt Foundation (see, e.g., Shiel, 2006).

18. Financial circumstances of nations (vis à vis other nations): Economic analyses of the performance of different nations have been assisted by the preparation after World War II of national accounts, on a relatively-standardized basis. (See Blejer and Ke-Young Chu, 1988; Guidotti and Kumar, 1991). The trading of government securities in international markets has created a market for assessments by credit rating agencies of the capacity of governments to repay debt (and possibly to assess currency risks).

Financial information encompassing the general government sector of all Australian governments (Commonwealth, state, territory and local government) is routinely compiled by the Australian Bureau of Statistics (see, e.g., ABS 2006). The information compiled includes a balance sheet, income statement, and statement of cash flows.

MAJOR ISSUES REGARDING IDENTIFICATION OF THE REPORTING ENTITY

The previous section reviewed judgments which might be made by prospective classes of users, and noted some evidence concerning the way information relevant to those judgments is currently presented—pointing to items of relevant information that could be provided by some form of consolidated statement. This section reviews several issues which emerge from the foregoing: (a) the scope of consolidated statements and the selection of supplementary consolidated statements; and (b) the boundaries of the government reporting entity (n.b., including whether consolidation should encompass various forms of trust accounts).

THE SCOPE OF PUBLIC SECTOR CONSOLIDATED STATEMENTS

If a single objective of preparing accrual accounting consolidated statements was specified, then it would be possible to develop a clear statement about the area of consolidation (i.e., the scope of the reporting entity). If consolidated statements are intended to serve several purposes, then the choice of the area of consolidation may involve trade-offs between the interests of different users of those reports.

For example, the area of consolidation could be determined by identifying what form of presentation provides information which is relevant to:

and so on. Inevitably, the presentation chosen will not provide information relevant to some judgments faced by some parties.

However, the above observations assume that a single set of consolidated statements are to be prepared—not (for example) multi-column consolidated statements prepared on differing bases. Much of the literature on the application of consolidation accounting to the private sector proceeded on the basis that only one presentation could be chosen (see Walker, 1976, 1978). Since that time, private sector financial reporting has evolved, so that company reports may present multi-column financial statements showing results on a conventional and equity accounted basis; components of the same data may be disaggregated in reports on the performance of segments. Correspondingly, it may be possible to prepare sets of consolidated statements in the public sector, possibly supplemented with accompanying schedules, to ensure the presentation of information deemed relevant to most of the judgments made by most users.

But what information is relevant to most of those judgments? Table 3 summarizes the material from Table 2 by listing only those judgments that may be made routinely.

It has been claimed that consolidated statements for whole of government should be regarded as general purpose financial statements. However, Tables 2 and 3 place those claims in perspective. While it has been asserted that general purpose financial reports are relevant to a range of users making a range of judgments, closer examination of many of the judgments identified in Table 2 indicates that the presentations of whole public sector consolidations (in the sense of reports relating to all Commonwealth, state or territory agencies) may not be routinely relevant to all categories of potential users. Table 2 also highlights the wide range of potential users of financial information which have quite specialized interests in financial data about the public sector.

One conclusion which might be drawn from this analysis is a variant of the familiar proposition that different forms of data are relevant for different purposes. The variant (or, rather, extension) is that different forms and arrangements of data are relevant for different purposes. Some of those forms or arrangements of data would be routinely relevant to decisions faced by some users; some data would be relevant only occasionally, or in particular situations.

Looking more closely at Table 3, several of those routine judgments (judgments #6 and #7) might be eliminated on the basis that an aggregated report might be less useful (and less relevant) to most stakeholders than a detailed analysis of the accounts of individual agencies. To illustrate: interest groups concerned with (say) public transport would obtain information relevant to their concerns from the financial statements of agencies actually providing those services rather than from a consolidated statement that also encompassed agencies providing public housing, or advice to primary producers.

Hence it is suggested that the forms of reporting which would provide financial information relevant to the routine judgments made by a range of prospective users are as set out in Table 4.

Even then, it may not be cost-effective to present so many sets of financial statements. For example, while it can be argued that a whole national public sector consolidation would provide relevant information for the purpose of comparing the financial circumstances of nations, it may not be considered worthwhile for individual national governments to prepare that type of report unless there was agreement that other nations would do the same. Similarly, while it can be argued that a state public sector and local government consolidation would convey relevant information for the purpose of comparing the performance of regional governments, it is likely that such high level comparisons could be undertaken by reference to some key financial aggregates or indicators (levels of liabilities, liabilities per capita or relative to Gross State Product, percentage of revenues allocated to new investment in infrastructure as opposed to recurrent spending, and so forth). That information may be more conveniently presented in the form of statistical data, as available in Australia from the ABS.

However, the focus here is on the criteria for identifying the scope of each of the main classes of consolidated statements that would produce information identified as routinely relevant. The foregoing analysis of users and potential uses, suggests that the optimal form of reporting by national, state and territory governments would take the form of ‘whole public sector’ reports, accompanied by consolidated statements encompassing both the general government and the non-financial public trading enterprise sectors.

It must be recognized that competing rationales for the determination of the scope of public sector consolidated statements have been presented. Some accounting standards dealing with whole of government consolidated statements have incorporated a test of control (e.g., Australian standard AAS 31, 1996, 1998; IPSAS 6, 2000). On the other hand, the U.S.A.'s GASB has emphasized that the appropriate tests for determining the scope of the government reporting entity should be based primarily on the notion of ‘financial accountability’ (GASB Statement 14, 1991). HM Treasury, for its part, largely avoided such difficulties by discussing the potential scope of consolidation by describing three categories of agencies or entities—central government, local authorities and public corporations¹⁵—and noting that ‘consolidation covering the whole public sector would therefore include of the order of 1,450 entities’ (1998, p. 24).

The significance of adopting one rather than another approach—determining boundaries by reference to control or accountability—can be examined by considering which test is likely to generate relevant information. But first it is worth recalling the rationale for the widespread use of the control test.

BOUNDARIES OF WHOLE PUBLIC SECTOR REPORTS

In the private sector, Walker (1978, pp. 261–65) observed that the techniques of consolidation accounting were developed in response to differing concerns. In the U.S.A., the earliest uses of consolidation accounting were intended to encourage public investment by demonstrating that sets of companies and related trusts comprised a major enterprise. Later, institutional factors reinforced the view that consolidated statements were desirable reports, including the levying of excess profits tax on groups of affiliated companies, thus requiring the use of a form of consolidated report. The aim of providing the means of assessing the credit-worthiness of holding companies or subsidiaries also came to be significant. The use of consolidated statements had become part of U.S. practice by the 1930s, though it was only in the late 1930s that any extended attempt was made to analyse in any detail the aims of consolidated reports.

In contrast, in the U.K. the advocacy of consolidated statements was a response to concerns about the limitations of financial statements in which inter-corporate investments were valued at cost (or derivatives of cost) and revenues from those investments were recognized only to the extent that dividends were received from investees. Consolidation accounting was regarded as one option, along with the separate presentation of the financial statements of holding and subsidiary companies, or what is now known as equity accounting (see Garnsey, 1923).

Again drawing on Walker (1978), during the 1920s major corporate collapses (the Insull and Kreuger groups in the U.S.A., the Royal Mail in the U.K.) appear to have reinforced arguments about the merits of consolidation accounting as a means of overcoming the distortions that could arise from inter-company transactions between controlled subsidiaries and the limitations of cost-based accounting and dividend-based revenue recognition in general.

However, the technical literature came to reflect differing views about the objectives of consolidation accounting—and hence about the relevant area of consolidation. Walker (1978) concluded that until the 1960s, British and Commonwealth regulations maintained different options for the presentation of group accounts, reflecting an ongoing view that consolidated reports were intended to amplify the accounts of a parent company. In contrast, while U.S. technical literature initially saw that percentage share ownership should determine the ambit of consolidation, texts began describing these reports as analogous to branch accounting, whereby a set of companies was treated ‘as if they were a single entity’. The analogy was then reinterpreted as a rationale for consolidation accounting: the purpose of these reports was said to be to present the affairs of a holding company and its subsidiaries as if they were a single entity.

Possibly for this reason, U.S. technical debate turned to consideration of how to identify whether a set of companies were acting in concert as a single entity, or a single economic entity. Possibly, the New Deal political objectives of regulating public utilities had some influence on accounting ideas, particularly in encouraging the suggestion that the reports should concern the performance of an economic entity. In that debate, some argued that a distinguishing feature of an economic entity was similarity of business activities—a rationale that would later be interpreted as excusing parents from consolidating their finance subsidiaries (a practice later seen as leading to the presentation of financial statements that were misleading—see, e.g., Heian and Theis, 1989). Others argued that the appropriate criterion for determining whether a set of companies should be regarded as a single entity was whether they were subject to common control, regardless of the nature of their activities. It is a matter of record that the notion that the ambit of consolidation should be determined by reference to control has prevailed on both sides of the Atlantic, and has been reflected in local and international accounting standards for private sector entities (e.g., Australian standards AAS 24, 1990; AASB 127, 2008; international standards IAS 3, 1976; IAS 27, 1989; IPSAS 6, 2000).

However, arguments that the same criteria should be translated to the public sector (arguments implicit in claims about the need for sector neutral accounting standards) can be seen as an illustration of confusing processes with objectives.

Arguably, one should start with a consideration of objectives, and then adopt criteria that are consistent with the selected objective.

For example, if the objective is to present a view of the financial circumstances of nations (judgment #18 above), then, as indicated in Table 3, it would be appropriate to consolidate Commonwealth, state and local governments—even though Commonwealth governments do not control the finances or the operations of state governments. This is consistent with the observations of the U.K.'s HM Treasury, which saw whole of government accounts as revealing information about a government's national fiscal policies—with the implication that public sector consolidated statements should encompass bodies that ‘exercise functions of a public nature’ or that are ‘entirely or substantially funded from public money’ (HM Treasury, 2008, Ch. 1 p. 3). As noted above, this would extend the ambit of consolidation to local and devolved governments. Whether there was a relationship of control between these different governments would not be relevant.

A second example: if the objective of public sector consolidations is to assess how a Commonwealth or state government has spent taxpayers' funds and any borrowings (judgment #8, above), then the focus of attention would be on the financial affairs of agencies that are part of the commonwealth or state government (such as government departments), together with other agencies that have been established by those governments.

A third example: if the primary focus of (some) potential users of financial information was on comparing the financial circumstances of regional governments (judgment #17 above) then the financial statements of state or territory governments should be consolidated with those of local governments. In Australia, local government has ‘a limited constitutional position . . .  being organized under state or territory legislation’ (ABS, 2006). In summary, local governments are not really governments, as they do not enact legislation and they are far from autonomous. State governments can dismiss councils and appoint administrators. They may also amalgamate councils into new entities. They can enact legislation compelling local councils (or country councils) to acquire or divest assets or to assume certain responsibilities. In some Australian jurisdictions (notably New South Wales) the revenues derived by local governments from property taxes are subject to rate pegging. In short, local governments are controlled by state governments. But the grounds for consolidating the financial statements of local and state governments would be derived from objectives, not a test of control.

Clearly, different criteria for identifying the scope of public sector consolidated statements would be appropriate, depending upon the objectives of those presentations.

Table 4 listed eight major options for the scope of public sector consolidated statements in order to provide information that is routinely relevant to major judgments likely to be made by potential users of aggregative financial information about the position and performance of different governments. Two of these (those relating to the ‘whole national public sector’ or the ‘whole state public sector’) reflect concerns with the overall performance of a national government vis à vis other nations, or of regional governments vis à vis other regional governments.

Another two of the eight major options reflect concerns with the accountability of elected governments (and hence countenance consolidation of the whole public sector at Commonwealth or state level).

The remaining four options are essentially supplementary reports covering the general government or non-financial PTE sectors of those governments.

Looking in more detail at the scope of whole public sector statements, there may be some agencies—trusts or special purpose entities—whose treatment may require careful consideration, and be debateable. Some examples follow:

School maintenance accounts: Government schools may be assisted in their activities by the fund-raising efforts of parents and others to contribute to the maintenance or improvement of physical infrastructure. Nominally ‘school maintenance accounts’ are controlled by those parents and citizen committees, but those resources are earmarked for application to the government-owned schools. Interpretation: once contributed, those resources are assets of the public sector, and the funds held in trust should be included in the accounts of the agency providing educational services and also included in public sector consolidated statements.¹⁶

Trusts holding unclaimed moneys: Statutory requirements are that certain unclaimed moneys are to be held in trust for a period, and if still unclaimed, then paid to the consolidated fund. Interpretation: those trusts are controlled by government, and government may theoretically have the power to seize those assets at an early date through the passage of legislation. Yet political realities are that this would be unlikely. While government is accountable for its stewardship of these funds, the funds are not available for government use until statutory periods expire. As previously noted, the trusts should not be consolidated.

Public guardianship trusts: Government has accepted responsibility to manage the financial affairs of persons who, because of intellectual or other incapacity, are unable to manage their own financial affairs. The assets of those persons are held in trust by a ‘public guardian’ or similarly named entity. Interpretation: While these trusts are controlled by government, and governments are accountable in a general sense for stewardship of the resources held in trust, the public sector has no beneficial interest in those resources. Those trusts should not be consolidated.

Trusts holding other assets: Trusts are often used as vehicles to manage public sector assets, ranging from areas that might be described as national parks, some heritage assets (such as historic buildings) and major playing fields (such as the Sydney Cricket Ground), to local sports fields or cemeteries. Use of trusts may be a convenient way to ensure local community or interest group participation in the management of the affairs of these assets. However, they remain custodians of public assets, and governments generally have the right to select and appoint trustees. Interpretation: These trusts should be consolidated.

Non-government organizations (NGOs) and public housing: A recent practice has been that state government agencies have assigned public housing assets to special purpose entities in the form of NGOs, which may be structured as cooperatives or as incorporated charities. The NGOs then rent the properties to persons on low incomes. Government agencies do not select members of the governing bodies of those NGOs. However, the process of assigning housing assets to NGOs is intended to ensure that services are provided to the community, and the government agencies may have rights to assume management control of the transferred assets. Interpretation: If the majority of the assets of those NGOs were acquired using resources contributed by taxpayers or from other charges, the financial statements of the NGOs should be consolidated to ensure that government stakeholders are informed about the assets and liabilities and financial performance of government. This interpretation is reinforced if the NGO's ongoing revenue stream is enhanced by statutory requirements for property developers to make contributions towards low income housing as a condition of development approval, whether those contributions come directly to the NGOs or are handled through local councils.

Entities established to manage compulsory insurance schemes: Governments may establish insurance schemes whereby levies are imposed on employers to meet commitments (established by statute) to injured workers. Those schemes may be managed by government-appointed boards. Actuarial assessments may be undertaken to assess whether resources in hand are held are adequate to meet emerging claims and to determine whether levies (commonly expressed as a percentage of employees' salaries) are adequate. Those levies come primarily from private sector employees. Interpretation: These schemes (however constituted) are controlled by government. Governments are also accountable for the conduct of these schemes, and the adequacy of their funding (and any government agencies that are participants in these schemes are exposed to increases in levies if resources held by the schemes are inadequate). Those entities should be consolidated.

Public sector superannuation schemes: Governments may have established pension schemes for public sector employees, and appointed persons to act as trustees (or directors of incorporated trustee entities). Schemes may broadly be classified as ‘accumulation’ or ‘defined benefit’ schemes. In both accumulation and defined benefit schemes, funds are held in trust on behalf of employees, pensioners (or their beneficiaries), and the public sector has no residual interest in trust assets. In defined benefit schemes, the government as employer has an obligation to make up any shortfall in order to meet the fund obligations; in some circumstances over-funding in a defined benefit scheme may revert to government control. Interpretation: While the liability or asset associated with under- or over-funding of defined benefit schemes should be reported as a liability or asset in the accounts of individual agencies (or in a public sector consolidated statement), inclusion of the significant assets and liabilities of superannuation funds in a public sector consolidated balance sheet would materially distort representations of a government's financial position, as those assets or liabilities cannot be applied for use by any public sector agencies to provide other (non-superannuation related) services to any section of the community.

State or Commonwealth funded universities: Many Australian universities were established by state legislation (and as such, their governing bodies include a significant but not a majority of government appointees). For decades, the bulk of their funding came from government grants—from the Commonwealth rather than the state governments under whose jurisdiction they fell. This division of responsibilities establish forms of checks and balances that may well have contributed to the operational independence of universities. However, a change in Commonwealth government policy led to a reduction in grants to universities, compelling them to place greater reliance on student fees (or investment income) to fund their operations. While those universities are not controlled by governments, and are no longer financially dependent on government (in the sense of deriving a majority of their revenues from that source), arguably the community sees universities as public institutions, and a resource of the public sector. They may not be controlled or financially dependent, but they remain a beneficiary of public funding and as such would be regarded as accountable to the broader community.

Table 5 summarizes the foregoing observations about whether the inclusion of certain agencies in public sector consolidated statements would generate information relevant to routine decision making. The headings used partly reflect the tests of control invoked by some accounting standards on public sector consolidation (e.g., AAS 31, 1998; IPSAS 6, 2000; AASB 1049, 2007), or are suggested by the specification of the government financial reporting entity such as in GASB Statement 14 (1991) in terms of tests of financial dependency and accountability.¹⁷

The Table 5 summary suggests that tests of control of a governing body are not the appropriate basis for determining the scope of public sector consolidated statements. Nor do notions of financial dependency via budgetary allocations constitute a necessary feature (as would be more obviously apparent if Table 5 had included public trading enterprises, rather than focus on the treatment of contestable cases). In any event, the notion of financial dependency is problematic. GFS tests for dependency relate to the majority of revenues; however, an agency receiving only, say, 10 per cent of its revenues from government budgetary allocations may be dependent on that funding for survival.

It seems incontestable that, as a minimum, the scope of a public sector consolidated statement for a state or nation government would encompass:

The foregoing analysis suggests that the following should be added:

These conclusions differ significantly from those reflected in prior studies that have recommended that the test of control be applied to determine the scope of public sector consolidated statements (IFAC, 2005, p. 25; see also AASB 1049, 2007). As noted above, this may well reflect a long tradition in accounting of focusing on processes (trying to find a suitable test for determining the ambit of consolidated statements) rather than focusing on objectives, and examining in some detail what information would be relevant to different users and uses. It might be noted that IFAC (2005) approached the issue from the perspective of harmonizing GAAP with GFS—rather than working from first principles as to what information was relevant to decision making.

Some of the above conclusions are similar to those reached by the U.S.A.'s GASB in its Statement 14 specification of the scope of the government ‘financial reporting entity’ (albeit in the context of preparing general purpose financial statements that provided a ‘combined statements overview’ of reports on sets of funds). The GASB's reasoning focused not on decision-usefulness but on financial accountability. Explanatory material in GASB 14 elaborated on the notion of financial accountability as follows:

Organizations that were controlled (in the sense that government could appoint the majority of the governing board) but did not meet the two subsidiary tests outlined in paragraph a above, were to be regarded as ‘related organizations’ but were not to be incorporated in the primary government's financial statements (see Granof, 1998, p. 494). Further, the standard elaborated on the concept of ‘financial benefit or burden’, as follows:

While the outcome of applying the tests embodied in GASB 14 are similar in some respects to the conclusions reached in this article, the GASB's analysis was based on considerations of accountability, not the relevance of data to judgments routinely made by stakeholders. It should not be surprising that, in some respects, those considerations intersect.

SUPPLEMENTARY CONSOLIDATED STATEMENTS

The foregoing Table 4 analysis suggested that some judgments potentially undertaken on a routine basis by users of public sector financial statements would be assisted by the provision of supplementary consolidated statements encompassing a jurisdiction's general government sector or non-financial public trading enterprise sector.

Recent accounting standards are consistent with this conclusion—although only in part. But first, some background. As was discussed in judgment #1 above, the lack of consistency between state governments in the identification of the budget reporting entity was the subject of some political controversy in Australia in 1991, leading to an agreement between states to work towards consistency in presentation. In practice, that involved some states continuing their own practices but only providing a reconciliation of reported budget projections with those that would have been provided had the budget encompassed all of the general government sector as defined in Government Finance Statistics.

In 2002 Australia's Financial Reporting Council¹⁸ noted that practices had developed whereby governments were producing financial statements (reporting actuals, not budget projections) based on generally accepted accounting principles while budget documents included financial results based on a GFS framework. Following submissions by governments in Australia and other stakeholders with an interest in public sector financial reporting, the Council directed the AASB to pursue as an urgent priority the harmonization of Government Finance Statistics (GFS) and Generally Accepted Accounting Principles (GAAP) reporting (FRC Bulletin 2002/5, 18 December 2002).

IFAC's Public Sector Accounting Standards Board, for its part, also adopted as a priority the convergence of accounting standards with ‘statistical bases of financial reporting’ and in 2004 the re-named International Public Sector Accounting Standards Board (IPSASB) initiated meetings of international organizations and regional government agencies or standard-setters and some national bodies to commence this process. IFAC (2005) analysed differences and proposed, inter alia, that general government sector information should be ‘allowed’ or ‘encouraged’ in whole of government general purposes financial statements, and that in these reports, ‘investments in controlled entities in other sectors’ be accounted for on a partial consolidated basis and measured at the proportional interest in net assets (p. 6). It was also suggested that disclosures about ‘other sectors’ should be allowed or encouraged.

The substantive outcome of these plans about reconciliation of standards-based accounting with GFS reporting was the issue by the AASB of an accounting standard AASB 1049, Financial Reporting by General Government Sectors of Government (September 2006), followed later by a revised version of AASB 1049, Whole of Government and General Government Financial Reporting (October 2007). The main elements of these standards concerned the specification of different sectors of government in terms of Australian interpretations of SNA 93, restrictions on the publication of a general government sector (GGS) consolidation before a whole of government consolidation was published, and the specification of how investments by GGS entities in other government agencies were to be valued and reported. The stated aim was to harmonize GFS and GAAP disclosures. In the event, AASB 1049 only required the provision of reconciliations of key fiscal aggregates with those that would be calculated in terms of the ABS's GFS Manual.

However, neither IFAC nor AASB (or other standard-setters) have yet to issue standards covering separate consolidation of PTEs—even though, as noted here, the case for the preparation of such as statement is equally compelling.

The case for presenting supplementary GGS and PTE sector reports was outlined in an earlier version of this article (Walker, 1995b). The AASB and IFAC publications have proceeded on the basis of harmonizing accounting standards with GFS (convergence) but have focused on reconciling concepts rather than arguing from considering what information would inform decision making. However, the main rationale for producing a GGS consolidated statement is that it could be read in conjunction with the formal budget previously introduced in parliament by the treasurer of the government of the day. Australian governments prepare budgets on an accrual basis, and while these documents incorporate data about budget results, these are unaudited estimates based on year-to-date data and projections compiled when budgets are finalized four to eight weeks before the end of the financial year. A consolidated statement on a GGS basis would report actual results (and those reports could be audited).¹⁹

It seems reasonable to rely on the UN or IMF definitions of the general government sector which are based on the extent of reliance of agencies upon funding from taxes. However, in light of the foregoing discussion, suggesting that tests of control are less relevant than other factors, it remains of concern that the IFAC (2005) paper has sought to influence national accounting by proposing that the convergence process should develop ‘common tests of control/boundary of the public sector and GGS’ (p. 7). Application of a test of control will exclude from the boundaries of the GGS those agencies that have been deliberately established so as to avoid government control of their governing bodies—an exclusion (it is argued here) that may not be appropriate.

The focus of much political or media attention is on annual government budgets—and on whether they represent a surplus or deficit. Standardizing the scope of the agencies to be encompassed by those budgets on the general government sector may avoid year-to-year manipulations of those results, and enable users of those reports to make considered judgments on the basis of information prepared on a consistent basis.

However, standardizing the area of consolidation may not be sufficient to ensure that budget projections or associated budget results are not subject to manipulation. A common form of manipulation arises from inter-sector transactions, or the push-down of unfunded mandates. Indeed, it is contended that it is necessary for an understanding of government finances to be able to identify the impact of the transfer of resources between the general government and other sectors.

To illustrate, the following are some Australian (N.S.W.) examples of inter-sector transactions. The scale of the transactions may not seem significant, relative to the scale of the funds reportedly located through the budget process in some countries (or, for that matter, in contemporary Australia). But they indicate how such transactions can be used in the public sector equivalent of private sector earnings management:

For that matter, the timing of transactions can also affect budget results:

Notably, some of these efforts to manage budget results could be avoided if such ‘special deposits accounts’ or ‘trusts’ were included in a GGS consolidation. However, the allocation of trusts to either the GGS or the non-financial public trading enterprise sector (NFPTE) may be problematic.

It might be argued that such a fund should be regarded as part of the (diminished) NFPTE sector on the ground that it was derived from the sale of shares in a public trading enterprise; however, investment returns were described as reducing public sector superannuation liabilities—some of which were attributable to employees of the agency that had been sold, and some to employees or pensioners of GGS agencies.

Further complexity arises with the establishment of other funds to support agencies that (contrary to the analysis presented above) have been categorized for the purpose of GFS as outside the scope of government—state universities:

If the intended use of a fund was to finance universities' investment in infrastructure, then the ‘purpose test’ would see it outside the GGS sector. But since it was established to reduce budgetary allocations of funds to universities in future years (and possibly allow for major grants to be undertaken selectively during the election cycle) then it could be argued that it was appropriately included in the GGS sector.

Mention might also be made of the financial public trading enterprise (FNPTE) sector—typically populated by one entity per jurisdiction—the entity charged with managing fund raising for government agencies and (in the Commonwealth arena) serving as a central bank. History suggests that the Reserve Bank of Australia has also been called upon to contribute to the Commonwealth budget via dividends— and that in some years, the payment of declared dividends has been held over to subsequent years (a practice that at the time would have deferred recognition of those dividends as a revenue of the Commonwealth's consolidated fund) (see ‘Reserve Bank payments to government’ in Reserve Bank of Australia, Annual Report 2005–06).

Similar observations may be made about the NFPTE sector, where reported operating surplus or deficits on an accrual basis may be manipulated by government grants, concessions or subsidies from a government's annual budget. A single example may suffice:

(If the ABS's definition of the GGS sector were applied, RailCorp would be regarded as part of the GGS sector since it is mainly financed by government. Yet in the 2006–07 N.S.W. public sector consolidated statements, RailCorp was treated as an NFPTE entity.²⁰)

Given the impact of these inter-sector transactions on reported results, the question arises: How best to ensure that stakeholders are adequately informed about the impact of these transactions (and are not misled by focusing on headline results)? One option may be to accompany suites of public sector consolidated statements with note disclosures about the nature and amount of those transactions as they occurred in the past financial year. Another may be the presentation of multi-column consolidating statements.

In the U.S.A. in the early part of the 1930s the newly formed Securities and Exchange Commission required public utility holding companies to accompany their consolidated statements with ‘consolidating statements’—described at the time as ‘formalized work sheets’ presenting the individual accounts of affiliates and showing the elimination of inter-company accounts (Childs, 1949, p. 257). This matter is further detailed in Walker (1978, pp. 249–50, 366). For public sector reporting purposes, consolidating statements could be less detailed, and confined to recording by line items, material transactions between GGS, NFPTE and the FPTE sectors.