Journal Issue
Share
Article

Promoting Fiscal Transparency the Complementary Roles of the Imf, Financial Markets and Civil Society

Author(s):
Murray Petrie
Published Date:
October 2003
Share
  • ShareShare
Show Summary Details

I. Introduction

The IMF’s Code of Good Practices on Fiscal Transparency was introduced in 1998 in the aftermath of the Mexican and Asian crises, and reflected an emerging consensus that a lack of transparency was an important factor contributing to those crises. The IMF, in conjunction with the World Bank, subsequently commenced a program assessing adherence to standards, codes, and principles in 112 areas, including the Fiscal Transparency Code, in the form of country reports known as ROSCs (Reports on the Observance of Standards and Codes), which are published with the authorities’ permission.

Greater openness by governments about their fiscal positions and intentions, other things being equal, is expected to improve their access to international capital markets, and to result in an earlier and smoother fiscal policy response to changing economic conditions, thereby reducing the incidence and severity of crises. In addition to providing a means for governments to assess and build their capacity for fiscal management, it is also anticipated that fiscal transparency may, over time, enhance oversight by civil society and the public of the conduct of fiscal management. In turn, both groups can provide positive reinforcement for countries to improve fiscal transparency. For this to occur, however, it is necessary that market participants and civil society are aware of, value, and utilize the information in fiscal ROSCs.3

This paper uses information from both surveys and interviews (done by the Fiscal Affairs Department in February and March 2003) to gain a better sense of both financial market and civil society views of fiscal ROSCs. The aims were to assess the initial impacts of the fiscal transparency initiative through these channels and to identify how fiscal ROSCs might be improved. Secondary objectives were to get feedback on and to publicize the fiscal ROSC initiative and related private sector and civil society activities. After discussing the findings of these surveys and interviews, this paper considers the relative roles of the IMF, financial markets, and civil society organizations (CSOs) in assessing and promoting fiscal transparency and how complementarity among their respective roles can be increased. It concludes by raising a number of possible ways in which the IMF’s fiscal transparency initiative might be made more effective.

II. The Role of the IMF

Through standard setting and the publication of assessments of observance of standards and codes (ROSCs), the IMF has taken the lead in promoting fiscal transparency. The IMF’s Code of Good Practices on Fiscal Transparency (the Code) represents a level of practices that all countries should be capable of achieving over the medium to longer term.4 In 1999 the IMF started a program of assessments of country adherence to the Code. The first fiscal ROSCs were experimental and very short, but since 2000 the coverage of fiscal ROSCs has expanded to cover all elements of the Code, and the format has been standardized. Section II of each fiscal ROSC contains a factual description of country practices against the Code, prepared by IMF staff on the basis of information provided by the country authorities – increasingly by completing a standard questionnaire prior to the ROSC mission. Section III of each ROSC contains IMF staff commentary on the level of transparency achieved, acknowledgement of recent reforms, and prioritized recommendations for improving transparency that take into account the existing level of transparency and institutional capacity.

The ROSC program is a voluntary one. Countries indicate an interest in participating in a fiscal ROSC assessment, often in response to a proposal by the IMF.5 Publication of the completed ROSC occurs with the consent of the country and after notification to the Executive Board of the IMF. As of July 2003, 59 fiscal ROSCs were completed and 54 were published on the IMF’s web site. The IMF intends to produce around 18-20 new fiscal ROSCs a year, and the target is to eventually cover most of the Fund’s membership. There has been broad participation across developed, developing, transition and emerging market countries, including a relatively high proportion of countries seeking access to international capital markets (over half of the sixty or so nonindustrial market access countries have chosen or plan to undertake a fiscal ROSC).

Box 1 describes key findings from an analysis of fiscal ROSCs completed by the Fund in 2003. Attempts are increasingly being made to integrate the findings into the Fund’s technical assistance, Article IV surveillance, and program design. In 2001, the Fund made standards one of the filters for prioritizing technical assistance requests. A recent Fund-wide review of the standards initiative found that fiscal ROSCs have in some instances helped to raise the profile of institutional weaknesses and specific concerns over fiscal data quality in the Fund’s surveillance discussions with the country authorities, as well as to help focus technical assistance.6

Box 1.Key Findings in Fiscal ROSCs

The Fiscal Affairs Department carried out an analysis of the completed fiscal ROSCs as at the end of October 2002. Some of the main findings were as follows:

  • ROSCs have identified a number of common problems that occur across a wide range of developing, emerging market and transition economies – in particular, problems of:

    • – poor fiscal data quality (such as poor coverage and consistency of fiscal data, weak internal control and audit functions, and unrealistic budgets);

    • – use of off-budget mechanisms (such as contingent liabilities and quasi-fiscal activities);

    • – lack of clarity in tax policy and excessive discretion in tax administration; and

    • – poor definition of inter-governmental fiscal relations.

  • Many of these issues are associated with a set of underlying institutional weaknesses that could lead to future fiscal or financial vulnerability – and these need to be addressed on a sustained basis.

  • However, often an immediate improvement in transparency can be achieved simply by publishing information already available within government.

  • Not surprisingly given the voluntary nature of ROSCs, most countries participating in a fiscal ROSC have undertaken or are undertaking significant reforms that increase transparency.

  • The EU-accession countries have made particularly impressive progress in increasing fiscal transparency.

  • Even among the industrial countries there are significant areas where fiscal transparency can be improved.

Source: Adapted from International Monetary Fund, 2003, Assessing and Promoting Fiscal Transparency - A Report on Progress, SM/03/86, Supplement 2, March 5, 2003. See http://www.imf.org/external/np/pdr/sac/2003/030503s2.htm

III. The Role of Financial Markets

As noted, financial markets are expected to play an important role in fostering the adoption of greater fiscal transparency. This role, however, depends on the awareness of financial analysts of fiscal transparency assessments, and on the value they attach to this information. More precisely, Schneider (2002, p. 10) has stated, in an article questioning the standards and codes initiative from a developing country perspective, that: “The key requirements for this [market incentives] to be effective would be (i) market familiarity with international standards (ii) their assessment of its relevance for assessments of market risk (iii) market access to information on compliance and the degree of compliance and (iv) use of information by the market in risk assessments.”

This section assesses the evidence with respect to these requirements. It first considers the potential reach of financial markets in the context of the increasing scope and scale of international capital flows. Second, market initiatives to disseminate information on adherence to the Code, and to independently assess countries against the Code are described. Third, evidence is presented from surveys and interviews on the level of financial market awareness of the fiscal transparency code and fiscal ROSCs, and of their relevance to investment decisions. Finally, this section concludes by considering the relative roles of the IMF and the private sector in assessing and promoting fiscal transparency, and suggests ways of enhancing the complementarity of their respective roles.

A. International Capital Flows and Fiscal Transparency

The large increase in recent decades in the number of countries borrowing on international capital markets, and the large increase in the volume of cross-border capital flows, has greatly expanded the potential reach of market incentives for increased transparency. For instance, only 30 countries had sovereign ratings in 1980, while by early 2003 111 countries had sovereign ratings. The most rapid growth occurred in the 1990s, during which the number of emerging market sovereigns with rated foreign currency debt increased almost seven times.7 It has been suggested that further growth may occur in the number of developing countries rated.8 There was also a well- documented large increase in the size of cross-border capital flows in the 1990s.

International capital flows comprise four main components: sovereign and corporate bonds; syndicated lending to sovereigns and corporations; investment in equities; and foreign direct investment (FDI). The existence of the “sovereign ceiling,” the linkage between sovereign ratings and regulatory and/or institutional investor restrictions on allowable investments, and the importance of country risk and governance issues to a wide range of international investment decisions mean that fiscal transparency can be relevant to some extent to all these market segments.9 However, the level of market interest in fiscal transparency is expected to be greatest in the sovereign bond and syndicated sovereign lending markets, because of the direct connection between a government’s cash flows and its ability to repay sovereign debt – particularly local currency debt. For instance, the methodologies used by the sovereign rating agencies incorporate a large role for fiscal data (including public debt data), fiscal management capacity, and quasi-fiscal and off-budget activities.10 The transparency of economic policy making is also a direct input into the rating agencies’ assessments of political risk. In turn greater interest in fiscal transparency is also likely with respect to emerging market countries – where markets typically find it harder to obtain information 11 – and amongst the sovereign rating agencies – because of their degree of specialization.12

B. Market-Based Initiatives to Disseminate Information on and to Assess Fiscal Transparency

Market awareness of, and use of information on fiscal transparency is evident from a number of private initiatives to disseminate information on fiscal transparency, and to independently assess the level of fiscal transparency in individual countries against the Code. An important example is the Estandardsforum, which was established in 2001 with the objective of increasing awareness in the private sector of the importance of compliance with standards to international risk assessments. It has been a subscription-based venture providing access to an internet database on countries’ observance across thirteen standards and codes, based on publicly available information.13 However, the information on the website is expected to be available without fee from mid-2003. The Estandardsforum could represent a significant development in bridging the gap between the information in fiscal ROSCs, and the quantitative rating of fiscal transparency that many market participants would like to see (see below). 14

Oxford Analytica provides Estandardsforum with an independent assessment of adherence to the Code in a number of countries for which there is no fiscal ROSC.15 Both IMF fiscal ROSCs and Oxford Analytica reports are then used by the Estandardsforum as a basis for assigning a rating on adherence to the Code for 83 countries.16 The rating is based on six categories (selected because they are considered to mirror the normal process a country follows when implementing standards and codes), in descending order: full compliance; compliance in progress; enacted; intent declared; no compliance; and no assessment available.17

Oxford Analytica has also been retained by the California Public Employees’ Retirement System (CalPERS) to produce assessments of fiscal transparency against the IMF Code in 27 emerging markets. While CalPERS found the Fiscal ROSCs useful, it considered their coverage of emerging markets to be proceeding at too slow a pace. CalPERS uses Oxford Analytica’s reports to generate a score for fiscal transparency, which it enters into an asset allocation model used to determine permissible emerging markets for CalPERS to invest in.18

In April 2003 CalPERs announced that it would give countries removed from its permissible investment list one full year to improve before they would take action to liquidate investments in these markets. Countries that have fallen below the threshold can obtain the information used to determine their rating, and CalPERs intends to support country efforts to achieve higher standards.

Finally, Credit Lyonnais Securities Asia (CLSA) produced a study in 2002 – available only to subscribers - that assessed how easy it is to understand the public accounts in each of twenty emerging market countries. The study entailed an assessment of each country against the IMF Code, the assignment of ratings by individual Code elements summed to an overall country rating, and the preparation of a league table.19

C. Market Awareness and Use of Fiscal Reports on Standards and Codes (ROSCs): Survey Results and Interviews

The official sector has conducted a number of surveys on the use of ROSCs of all types by financial markets. These surveys found an initially low, but growing level of awareness.20 Use of ROSCs was greatest amongst US-based institutions with large global networks, and least amongst Japanese institutions. European institutions’ use of ROSCs was mixed, but around three quarters of the twenty three respondents were aware of ROSCs, and over half had used ROSCs directly in risk assessment. Fiscal ROSCs were cited as being particularly useful to institutions that focus on sovereign risk and macroeconomic conditions. Market feedback has consistently been that ROSCs in general could be improved by being shorter, more up to date, and covering more countries.

In order to assess market views of fiscal ROSCs more precisely, the IMF sent a survey to the three major sovereign rating agencies; to about ten major US-based international banks and investment banks; to around eighty institutions selected by the Institute for International Finance (IIF) from its global membership; and to Oxford Analytica, the Estandardsforum, and the Economist Intelligence Unit. Interviews were also held in New York and England to discuss the issues in greater detail (a list of those interviewed is contained in Appendix III).

The largest number of responses was received from the rating agencies. The surveys revealed a high level of use of fiscal ROSCs, and that fiscal ROSCs are having some impact on sovereign risk assessment. Box 2 contains the key findings from the survey of rating agency analysts - including views on fiscal data quality - while Appendix I contains a description of the survey methodology and a summary of the quantitative results, and Appendix II contains a synthesis of the qualitative responses.

From interviews with rating agency staff, it is apparent that both Standard and Poors and Fitch have built the use of fiscal ROSCs into their standard operating procedures. For instance, Standard and Poors analysts consult the fiscal ROSC (where there is one) either before a country visit or prior to the rating committee meeting.21 Moody’s, on the other hand, leaves it up to the individual analysts to decide whether to look at a fiscal ROSC.22

The interviews also revealed that rating analysts and US-based investment banks look to ROSCs for a clear indication of whether the general government sector is defined in accordance with international standards, and therefore whether the government’s fiscal data are comparable with other countries. In addition, both groups are particularly looking to fiscal ROSCs for information on off-budget activities, contingent liabilities, and quasi-fiscal activities, quantified wherever possible. Ideally they would like ROSCs to indicate the true underlying fiscal deficit or public debt in each country. About half of all survey respondents strongly agreed that fiscal ROSCs should contain more quantitative data.23

Amongst the selected major US-based international banks and investment banks, the interviews revealed an increasing interest in information on the integrity of fiscal data. This is in part due to the influence of the new Basel Capital Accord, which will entail major international banks developing their own internal country risk rating methodologies. Amongst this group of institutions, some want just a single quantitative rating of fiscal transparency as an input to management of corporate risk, while others find the complete fiscal ROSC useful for countries where they have major credit exposures.

The IIF survey was intended to test the level of awareness and use of fiscal ROSCs outside the US, and amongst a broader range of investor types. Unfortunately the response rate was too low to provide any meaningful data. This is an aspect that requires further exploration.

Box 2.Rating Agency Survey: Key Findings

Twenty one analysts in the sovereign risk units of Standard and Poors, Moody’s, and Fitch responded to the survey (response rate 50%). The main findings were:

  • 71 percent had read a fiscal ROSC or some part of a fiscal ROSC.

  • Of those, 93 percent said they had a better background understanding of the country’s fiscal position and fiscal risks as a result of reading a fiscal ROSC. About half of them had used information in the fiscal ROSC as a direct input into their rating assessment.

  • In response to a question on how significant an impact a fiscal ROSC had had on a country risk assessment, on a scale of 1-5 (1 indicating no impact, 5 indicating a major impact) the mean response was 2.3. Six respondents said a fiscal ROSC had a major impact (5) on the assessment and six others responded with a 4.

  • Analysts rated the most important features of fiscal ROSCs as ease of accessibility; consistency and comparability across countries; a source of information on fiscal management institutions and practices; and inclusion of IMF staff recommendations on reforms needed. Written comments particularly reinforced the desire for international comparability.

  • Analysts would most like to see more quantitative data in fiscal ROSCs, and, to a lesser extent, a quantitative rating of fiscal transparency. They do not consider fiscal ROSCs to be too long, to lack candor, or to be up-dated too infrequently.

  • Analysts consider fiscal data on the governments they monitor to be adequate, and to have improved somewhat over the last 1-2 years. A number mentioned improved accessibility through posting of data on government web sites.

  • Asked whether a more widespread use of accrual basis reporting would improve their ability to assess sovereign risk, on a scale of 1-5 (1 indicating no impact, 5 indicating major impact) the mean response was 3.4. Many respondents were more concerned about having data that is internationally comparable than with which accounting basis is used. A number also stressed the need for full reconciliation between accrual and cash numbers

  • Analysts place the highest priority on having data on off-budget expenditures, contingent liabilities, public debt net of financial assets, quasi-fiscal activities, the medium term fiscal outlook, and the sensitivity of the budget to changes in macroeconomic assumptions. Least interest was shown in non-financial performance information. 1 Only 33 percent of respondents were aware of private sector assessments of fiscal transparency. Most of those who answered the question comparing fiscal ROSCs to private sector assessments considered them to be complementary to fiscal ROSCs, rather than potential substitutes.

  • Around 90 percent of respondents indicated they may make more use in future of published material on fiscal transparency produced by the IMF or others.

1/ In interviews with some ratings analysts, it was indicated they look particularly to fiscal ROSCs for information on off-budget and quasi-fiscal activities and on contingent liabilities, while they rely on their discussions with the authorities for information on the medium term fiscal outlook and macrofiscal sensitivity.

Finally, as expected, given they are a direct input into their own assessments of fiscal transparency, the level of usage of fiscal ROSCs by Oxford Analytica and the Estandardsforum is very high, and the fiscal ROSCs have a major impact on their assessments.

D. The Relative Roles of the Fund and Financial Markets

According to the surveys and interviews with market participants, the range and frequency of Fund staff contact with the national authorities are seen as a major advantage in assessing country adherence to the Code. However, countries choose whether to participate in an assessment, and subsequently decide whether to agree to publication of the assessment. While a high percentage (around 90%) of completed fiscal ROSCs have been published, this generally occurs following exchanges with the authorities over the wording of the ROSC (unlike IMF staff reports, ROSCs are shared with the authorities in draft). Although the mean survey response on whether fiscal ROSCs lack candor was under three (mild disagreement with the statement that ROSCs lack candor), feedback from some market participants suggested there is a perceived lack of frankness and independence in fiscal ROSCs due to the fact the Fund is evaluating its own member countries. Some participants saw this as a significant issue. Some also observed that there are potential tensions from combining the standards assessment function with the other operational responsibilities of the Fund.

The Fund is strengthening internal guidelines to make it clearer that the judgments expressed in a ROSC are those of the staff, while the views of the authorities should be reflected in an accompanying right of reply. A standard disclosure statement should be added to the cover page of each fiscal ROSC stating the publication policy (publication occurs with the agreement of the country authorities) and that the Staff Commentary Section represents the views of IMF staff only.24

The Fund’s fiscal transparency assessments are qualitative in nature. The IMF’s Executive Board has explicitly decided that the Fund will not provide a quantitative assessment or rating of country transparency, because of concern about the misleading nature of simple pass/fail assessments. Survey responses strongly suggest that many (though by no means all) private sector participants want a quantitative rating (41% of all respondents strongly agreed that fiscal ROSCs should contain a quantitative rating of fiscal transparency).

These issues raise the question of whether the Fund’s assessment role could be carried out by non-official sector entities not subject to the same constraints as the Fund. Assessing fiscal transparency entails to a large extent evaluating information that is already in the public domain. Indeed the Estandardsforum explicitly uses only publicly available information in its assessments – in part in the hope that this will promote the release of information into the public domain. There is already a private market in producing fiscal transparency reports assessing most emerging market countries against the Code (with annual up-dates). These are currently financed by the management fees charged by a mutual fund, but in principle might instead be financed by subscribers to the advisory services provided by an investment bank25 or by the governments themselves that issue international bonds.26

One issue is the extent to which such private market assessments could potentially substitute fully for the IMF’s fiscal ROSCs. There are a number of considerations here:

  • First, it would seem likely that any sustained private market in fiscal transparency assessments would be limited to emerging market countries, or to additional countries of particular interest to international investors, rather than to all member countries of the IMF.

  • Secondly, if the assessments are financed privately they are likely to be treated (at least to some extent) as proprietary information and not be available in the public domain.27

  • Thirdly, there is an issue about the level of resources that could be profitably devoted to producing assessments, and therefore the depth and quality of the assessments.28 For instance, the Oxford Analytica fiscal transparency reports are produced with considerably less time input, with less easy access to information, and are considerably shorter than fiscal ROSCs. While it is not obvious that the additional resources invested in an IMF ROSC, and the additional information it contains, represent value added, survey and interview responses suggests the additional depth and breadth of a fiscal ROSC is regarded positively – and that private sector assessments are viewed as complements to, rather than substitutes for ROSCs.

  • Finally, ratings agencies and international investment banks that might otherwise be well positioned to produce such assessments have other commercial interests with governments, and may not wish to complicate those relationships by publishing fiscal transparency assessments that raise sensitive governance issues. Investment banks might at times also face conflicts of interest that could constrain their independence.

There would therefore appear to be little risk that fiscal ROSCs might crowd-out a full private market in fiscal transparency assessments - at least for the foreseeable future. Fiscal ROSCs have the characteristics of an international public good. They aim to meet the needs of a diverse range of international and domestic financial market users (as well as the needs of legislatures, civil society and so on), and there are economies of scale and scope in their production that the IMF is well placed to exploit.

E. Increasing Complementarity Between IMF and Private Sector Activities

The most interesting questions therefore involve the boundaries between fiscal ROSCs and related private initiatives, and whether complementarity between the approaches can be enhanced. Given the qualitative nature of fiscal ROSCs, a key issue is the extent to which the private sector might play a role in bridging the gap to the single quantitative transparency rating that many outside the official sector want. The Estandardsforum, as noted, is disseminating country fiscal transparency ratings, and the anticipated civil society funding of the Estandardsforum may see the country coverage greatly expanded, further extending the impact.

The fact that both participation in and publication of the ROSCs are voluntary also creates scope for private sector assessments to add value with respect to those countries that have yet to agree to publication of a fiscal ROSC. Most value could be added here for any systemically-important countries that have yet to be assessed; for countries active on international capital markets; and for those regions where there is a relative lack of participation.

Private sector up-dates of fiscal transparency assessments could also fill an important gap given the limited number of fiscal ROSC up-dates produced to date by the Fund – and the likely resource constraint on the Fund producing as many up-dates as the market and others would like.29 Oxford Analytica is, for instance, currently producing annual up-dates of its fiscal transparency assessments. The Estandardsforum also posts information under the heading “Positive Developments”, indicating whether there have been any news items from its Weekly Reports over the past two quarters that indicate positive progress by a country in implementing standards and codes.

As noted, markets are looking for more quantitative data in fiscal ROSCs, especially on off-budget and quasi-fiscal activities, and contingent liabilities. ROSCs could do a better job on occasions of signaling whether a non-observance of a Code practice is a minor technical breach or of potentially substantive importance. But to actually quantify off-budget activities would mean a fundamental change in the nature of the fiscal ROSC. It would considerably add to the time and complexity involved in completing a ROSC. In the absence of any response by the authorities to develop and publish quantitative estimates, private sector analysts with responsibility for detailed fiscal analysis – such as ratings analysts – are well placed to take these issues up during country visits in an attempt to put numbers on any gaps in their knowledge. Similarly, IMF area departments are in a better position to pursue these issues with the authorities than a ROSC mission. If the Article IV report is subsequently published, markets will be informed through that route.

Short of a quantitative rating, one or two survey respondents said they would welcome information in each ROSC explicitly comparing fiscal transparency across countries. This would be a significant additional requirement. The Fund has, however, recently completed an exercise analyzing the findings from fiscal ROSCs – see IMF, 2003d. Annex II of that report contains tables providing simple cross-country comparative assessments of fiscal transparency observations drawn from fiscal ROSCs. Further periodic assessments could also be worthwhile. In addition, consideration could be given to posting a periodic Fiscal Transparency Update on the web site that provides information on fiscal ROSCs published in the previous period, and pulls together information relevant to fiscal transparency in individual countries contained in Fund publications (such as Article IV reports, etc).

Complementarity of efforts between the Fund and private sectors depends on the level of awareness of the availability and content of fiscal ROSCs. While there seems to be considerable (although still uneven) use of fiscal ROSCs by sovereign rating analysts, and by some in the US international investment community, the evidence suggests the level of knowledge is very limited in the broader international investment community. There is a similar issue with respect to civil society. A number of suggested approaches the Fund could take to increase awareness are outlined in the conclusion of this paper.

There is scope for financial markets to more proactively encourage countries to signal their level of adherence to the fiscal transparency code. For example, sovereign rating agencies could explicitly refer to fiscal transparency issues in their rating comments on individual countries. They could also discuss fiscal transparency issues and the linkages to sovereign risk in their general publications on sovereign rating methodologies. Underwriters could recommend to the authorities that they include references to fiscal ROSC findings in their offering memoranda when underwriting public sector issues from emerging market economies.30

Finally, until very recently, difficulties in obtaining measures of fiscal transparency were a major obstacle to research. The codification of what constitutes fiscal transparency, and the public availability of fiscal ROSCs and published cross-country analyses of fiscal ROSC findings, is now generating a growing body of comparable cross-country data on fiscal transparency. In addition, private sector quantitative ratings of country adherence to the Code, and civil society ratings of budget transparency (see Section IV) are also beginning to emerge. Frameworks that combine fiscal transparency assessments with assessments of fiscal sustainability and fiscal vulnerability, or with fiscal management capacity, are available.31 These developments promise to make possible a more rigorous testing of the proposition that fiscal transparency impacts on fiscal performance, on high quality growth, poverty reduction and other goals. They should also make it possible to start to identify which aspects of fiscal transparency have the largest effects. Such research information could strengthen the interests of markets in information on country adherence to the Code, and reinforce the incentives on countries to signal a credible commitment to greater transparency.

IV. The Role of Civil Society

In addition to improving the interactions between governments and financial markets, the preamble to the original Fiscal Transparency Code also referred to a somewhat broader set of objectives, stressing the contribution fiscal transparency would make to the cause of good governance. Fiscal transparency “should lead to better informed public debate about the design and results of fiscal policy, make governments more accountable for the implementation of fiscal policy, and thereby strengthen credibility and public understanding of macroeconomic policies and choices.” (IMF Survey 1998, p. 122).

In practice, the Fund and institutions such as the Financial Stability Forum (FSF) and the Institute of International Finance (IIF) have tended to focus on the interactions between fiscal transparency and financial markets – and in particular on international financial markets.32 In the Fund’s case, this reflects its core macroeconomic mandate. The potential role of civil society in promoting fiscal transparency has been the subject of relatively little attention – although it would seem to offer considerable longer term potential for bringing about a sustainable strengthening of domestic institutions for more accountable and effective fiscal management.

This section discusses recent initiatives by civil society to promote greater transparency and accountability in government budgeting at the country level, and the scope for greater complementarity between the efforts of civil society and the Fund.

A. The International Budget Project

In 1997 the Center for Budget and Policy Priorities (CBPP), a Washington DC-based NGO specializing in applied budget analysis, established the International Budget Project (IBP). The IBP aims to assist civil society organizations (CSOs) and researchers in developing countries and new democracies in their efforts both to analyze budget policies and to improve budget processes and institutions. The IBP involves an informal network of CSOs currently actively working in about twenty countries, and having indirect engagement with groups and individuals from another sixty or so countries. 33

The IBP has focused on budget transparency and participation because it recognized early on that poor access to information was a major barrier in many countries to civil society’s efforts to influence public debate over government spending. It also recognized that promoting open budgets can attract broad support, including domestic business interests, international organizations, and others.

B. Country Budget Transparency Studies

In 1999 the IBP collaborated with the Institute for Democracy in South Africa (Idasa) in a pilot budget transparency and participation study of South Africa. The method used in the study drew particularly on the IMF Code, but went beyond it by introducing public participation in the budget process as a core element in the definition of fiscal and budget transparency. “…it was felt that the technical and strategic advocacy benefits of using codified international standards and published prior research justified their use.” (Folscher, 2002, p. 4). The Idasa study compiled a three-step rating (weak, moderate, good) of transparency and participation in South Africa in each of five categories, and an overall country rating.34

In December 2001 five groups in Latin America published the results of budget transparency studies of Argentina, Brazil, Chile, Mexico, and Peru. These groups used a somewhat different methodology, involving a detailed expert assessment and rating of the legal frameworks for budget management in each country, and a budget transparency index that rated each country on a scale of 1-10. The index was based on perceptions of budget transparency obtained from surveys of legislators, civil society, and other key users of the budget outside government.35 The reports were released through a simultaneous press conference in each country, and subsequently through a series of briefings in Washington DC.

The Africa Budget Project, led by Idasa, completed a five-country comparative study of budget transparency and participation in 2002. The study covered Kenya, Ghana, Zambia, Nigeria and South Africa, and contained detailed country case studies. In addition, the study rated each country in terms of three levels (weak, moderate, good) against transparency and participation criteria, and the adequacy of their legal frameworks.36

A stand-alone country study of fiscal transparency in Poland was completed by the Gdansk Institute for Market Economics in 2001. Interestingly, this study explicitly assessed Poland against the Fiscal Transparency Code. It was produced independently of the fiscal ROSC for Poland, which had been published earlier in the same year. The study noted that: “In our opinion, the state of openness and transparency of public finance in Poland is slightly worse than according to the IMF [in the ROSC].” 37

A small number of CSO budget transparency studies have also been completed at the sub-national level. For example, Strategia completed a report card measuring budget transparency at the local and regional government level in several Russian provinces; and the DC Fiscal Policy Institute published a budget transparency scorecard for the District of Columbia in the United States.38

Finally, there has also been some civil society interest in fiscal transparency as an important component of a country’s national integrity system, and in the use of the Code as a high level diagnostic tool in anti-corruption work. For instance, the Transparency International (TI) Source Book contains a description of the Code, highlighting those elements that bear directly on the incidence and severity of corruption.39 TI National Chapters are reportedly taking an increased interest in campaigning for open budgets.40

C. Civil Society Awareness and Use of Fiscal ROSCs

It is clear that the IMF Fiscal Transparency Code has had a significant impact on the efforts by CSOs to develop and apply methodologies that measure budget transparency. A number of CSOs have also drawn on the Fiscal Transparency Manual in their technical analysis work. Notwithstanding this, however, a recent IMF survey found that knowledge of, and use of fiscal ROSCs is limited amongst civil society – although it is difficult to be sure due to the low survey response. Even amongst CSOs participating in the IBP, knowledge of fiscal ROSCs appears to be low – despite the fact that there are links on the IBP web site to most fiscal ROSCs. See Box 3 for a summary of civil society awareness and use of the fiscal Code and ROSCs.

The low level of awareness is not surprising. Both the Fund’s fiscal transparency initiative and international civil society budget transparency initiatives are relatively recent. There is very little overlap to date between the countries covered both by budget transparency studies and fiscal ROSCs – only Poland, Mexico and Brazil are in both groups, and in the latter two cases the fiscal ROSC was published only after the NGO studies were completed.41 The Fund has not actively publicized fiscal ROSCs amongst civil society groups, the content and language in fiscal ROSCs is highly technical, and the Fund may be regarded as an unlikely source of such information.

D. Increasing Complementarity Between the Efforts of Civil Society and the IMF

The IBP is currently assessing the scope for a significant expansion of international budget transparency efforts by CSOs. A common core methodology is being developed to measure budget transparency consistently in different countries and regions, and thereby to promote efforts to influence budget outcomes and social justice around the world. The aim is to develop a budget transparency scorecard that allows consistent comparisons across countries and through time. 42 Consideration is being given to launching a formal International Campaign for Open Budgets.

As with private market initiatives, it seems very unlikely that civil society studies could substitute fully for fiscal ROSCs – although for different reasons. First, while civil society studies are publicly available, and there is considerable commonality of interests between fiscal transparency and budget transparency, CSOs have additional objectives and different areas of focus. They are therefore very unlikely to want to assess countries against all the elements in the Code. Second, civil society would face formidable resource constraints in terms of an exercise of this magnitude. Third, it would be difficult to coordinate the studies of independent CSOs across all countries and regions, although as noted above the efforts of the IBP to develop a scorecard are intended to address some of these challenges.

Box 3.Summary of Views of Civil Society

Two separate surveys were distributed: one to the groups participating in the CBPP’s International Budget Project, the other to a broader group of international development CSOs.

The main findings from the survey of international development CSOs were:

  • Of the ten respondents, three were aware of the Fiscal Transparency Code, two were aware of fiscal ROSCs, and one had read and used part of a fiscal ROSC.

  • Of the ten respondents, three were aware of NGO budget transparency studies, and two had read some part of a study.

  • All ten respondents indicated they anticipated making more use in future of information published by the IMF on fiscal transparency.

  • Presented with possible options for the Fund to more actively disseminate information on fiscal transparency, most support was expressed for issuing a periodic Press Notice indicating fiscal ROSCs completed in the previous period; and issuing a Press Notice in-country when each fiscal ROSC is published.

The main findings of the CBPP survey were:

  • Of the five respondents, four were aware of the Fiscal Transparency Code and fiscal ROSCs, and one had read or used part of a ROSC.

  • Respondents reported that there has been little improvement in budget and expenditure data published by the government and that it is insufficient for their needs.

  • Asked about how the IMF could better publicize and disseminate its fiscal transparency initiative, most interest (4.4 average response out of 5) was shown in presentations to civil society groups and issuance of in-country press notices.

There are, however, a number of areas in which the Fund’s efforts and those of civil society are complementary, and could be more so in future:

  • While the Code and fiscal ROSCs take a broad view across all fiscal activities and focus more on the macroeconomic level, the budget transparency studies focus more narrowly on the spending side of the budget, where they have the capacity to go into detail on transparency and accountability at the sectoral and program level. In that sense they can drill down more deeply than fiscal ROSCs into the micro level issues that are more directly related to poverty alleviation – such as the design and monitoring of Poverty Reduction Strategies, and whether they are adequately integrated into national budgets.

  • Reflecting the mandate of the Fund, the Code focuses on national government transparency, and covers some, but not all the transparency issues at the sub-national level. Important issues of fiscal transparency arise in many countries with respect to the broader activities of sub-national government, especially given the move in many countries to devolve responsibility for public spending. The CSO sector may have a potentially valuable complementary role to play in assessing budget transparency and accountability of sub-national governments – and many of the principles in the Code can be applied to sub-national governments.

  • By assessing also the opportunities for, and actual level of, public participation in fiscal policy debates and budget allocation decisions, CSOs can help to assess the level of legitimacy of current fiscal policies, which is a factor in assessing the sustainability of fiscal policy.

  • The cross-country rankings being produced by CSOs may be a useful complement to the qualitative information in fiscal ROSCs.

  • There appears to be considerable scope, as the number both of CSO budget transparency studies and fiscal ROSCs expands, for civil society to leverage off fiscal ROSCs, both in terms of analytical and advocacy activities. This will require a greater level of awareness of the availability and usefulness of fiscal ROSCs amongst civil society. (Suggestions for increasing the level of awareness are made in the conclusion).

  • Finally, civil society budget transparency studies could, in turn, prove to be a very useful complement to, and cross-check on the quality of fiscal ROSCs. CSOs will often be independent of government, and they will often possess a comparative advantage, in terms of intimate knowledge of how interactions between government and non-government sectors operate in practice in their country. This is a key input into assessing fiscal transparency in areas such as tax administration, procurement, expenditure arrears, regulation, and the availability in practice of fiscal information to the general public. More generally, they will also be well placed to observe the extent of congruence or mismatch between formal legal requirements and actual practices on the ground – a key difficulty in assessing fiscal transparency – and to advocate improvements to both laws and practices.

V. Conclusion

The IMF’s endorsement of the Fiscal Transparency Code and its program of fiscal ROSCs have the characteristics of an international public good. Fiscal ROSCs inform a diverse range of users–legislatures, civil society, the business community and the wider public within individual countries; academics and researchers; and international investment, civil society, and donor organizations. They support the growing international demand for more comprehensive fiscal data and information on fiscal policies and the effectiveness of fiscal management; and they encourage countries to adopt more transparent practices and to produce and publish higher – quality fiscal data. The fiscal ROSC program may even have an impact on countries that have not chosen to undertake a fiscal ROSC by affecting the authorities’ willingness to amend policies or practices in line with the Code.

In addition, the IMF is well placed to exploit economies of scale and scope in ROSC production – although there are some concerns about the objectivity and independence of fiscal ROSCs. Given the dual role of the IMF in this area as both standard setter and assessor, it is important in the future to have an independent review of the Code itself, including wide consultation and opportunity for input on the detailed requirements it contains.

Financial markets are expected to play a key role in fostering the adoption of greater fiscal transparency. This depends on the awareness of private sector analysts of fiscal transparency assessments and on the value they attach to this information. Survey responses and interviews show considerable use of fiscal ROSCs by sovereign rating analysts: they are being progressively incorporated into their standard operating procedures, and they are having some impact on sovereign risk assessment. Some in the US international investment community are also actively using fiscal ROSCs as an input into their credit assessment, investment allocation, and corporate risk management decisions. Market demand for information on fiscal transparency is also evident from the production of independent assessments of fiscal transparency, from the assignment of country ratings, and from initiatives to actively disseminate information on country compliance.

Financial markets and civil society can further leverage the impact of fiscal ROSCs in a number of ways. Rating agencies and bond underwriters could more proactively signal the importance of fiscal transparency and of participation in fiscal ROSCs. As the overlap between countries with both a fiscal ROSC and a budget transparency assessment increases, civil society groups may more actively draw on fiscal ROSC findings in their technical and advocacy activities. In turn, the increasing availability of budget transparency and participation assessments has the potential to complement the fiscal ROSC program in a number of respects (such as adding more detail on transparency of public spending at the sectoral and program levels, and in-depth assessments of subnational government).

Although there is a lack of evidence, it appears that awareness of fiscal ROSCs is relatively limited outside the major sovereign rating agencies and major U.S.-based financial institutions. It also seems likely that the level of awareness of fiscal ROSCs is very low among broader civil society groups and that applied budget analysis CSOs have to date made very little use of ROSCs —although they have used and adapted the Code. More focus on outreach to both civil society and the private sector could help promote synergies that will reinforce country efforts to become more fiscally transparent.

There are a number of possible approaches the IMF might consider to increase both the awareness and usefulness of fiscal ROSCs:

  • Fiscal ROSCs should

    • differentiate more clearly between significant and minor nonobservances of the Code;

    • pay particular attention to describing the extent and quality of publicly available information on off-budget activities, contingent liabilities, and quasi-fiscal activities. Where quantitative assessments of these have been produced (for example, in published IMF or World Bank reports), these should be cited;

    • provide a clear assessment of the extent to which a country’s fiscal data are consistent with recognized international standards.

    • provide better signposting of the contents of the ROSC to ease the task of infrequent readers or those interested only in the Executive Summary or the IMF staff commentary;

    • disclose the publication policy on the cover page; and

    • provide a contact point for inquiries.

  • Fiscal ROSC missions should, with the permission of the authorities, routinely meet in-country with legislative staff and officials, and with representatives of the domestic business community, civil society organizations, and academics and researchers in the course of completing their fiscal transparency assessment.

  • Greater outreach efforts would be worthwhile including such activities as:

    • Regional seminars for officials or financial market representatives and forums with NGOs could be offered.

    • Presentations on the Code could be given to legislators and civil society groups especially in PRSP countries.

    • The publication of each completed ROSC could be accompanied by a plain-language press release aimed at nonspecialists, which would be disseminated within the country as well as being posted on the IMF’s website.

    • IMF missions and resident representatives could be more proactive in flagging key issues identified in fiscal ROSCs in their presentations and press conferences.43

    • ROSCs could be translated into the national language for in-country dissemination.

    • An e-mail notification service could be offered to those interested in being informed when a new fiscal ROSC or ROSC update is published.

  • The information on the SDDS Bulletin Board on the fiscal sector could be supplemented, for example, with the addition of lines indicating whether contingent liabilities are disclosed and whether a fiscal ROSC is available.

  • Consideration could be given to posting a periodic Fiscal Transparency Update on the web site that provides information on fiscal ROSCs published in the previous period and pulls together information on developments relevant to fiscal transparency in individual countries contained in recent IMF publications.

  • The IMF could publish further periodic analyses of the findings of fiscal ROSCs, including cross-country comparisons.44 It could also make its fiscal ROSC database and search tools available on its external website to make it easier for users to search by issue or group of countries.

  • Last but not least, greater efforts by the IMF to produce and support new research on the impacts of fiscal transparency on fiscal and financial market performance, and to disseminate the results of the research widely could be of major benefit.

APPENDIX I

I. Financial Institutions Survey: Methodology and Summary of Quantatative Results

Separate surveys were developed for financial institutions, for budget analysis CSO, and for broader civil society groups. They were sent by email, and respondents were asked to complete the surveys electronically and return them by email. In two cases a hard copy of the survey was handed to a respondent; in about four cases a hard copy response was returned.

Financial Institutions Survey

The survey was intended to explore the level of awareness and use of fiscal ROSCs, what features and information are most valued, their perceived strengths and weaknesses, and to elicit views on how ROSCs could be made more useful. The survey also aimed to assess the level of awareness of private sector (and civil society) fiscal transparency initiatives, and views on different aspects of fiscal data quality.

It was sent to the three major sovereign bond rating agencies: Standard and Poors, Moody’s Investor Services, and FitchRatings. In the case of Standard and Poors and Fitch, a manager within the rating unit sent the survey to all the other members of the rating unit (except for senior management). Moody’s discussed the survey at a meeting of its ratings unit, and four analysts volunteered to complete the survey, three of whom actually responded. In an attempt to obtain a more representative sample, the IMF also sent the survey direct to two other Moody’s analysts selected at random, one of whom responded. A Moody’s analyst also sent the survey to four other analysts, none of whom responded.

The survey was also sent to individuals in US-based investment banks and financial institutions who were known, from a previous IMF outreach exercise, to be generally aware of ROSCs, in order to obtain feedback specifically on how and to what extent they were using fiscal ROSCs, and their views on the quality of fiscal ROSCs.

The survey was also sent to members of the Institute for International Finance (IIF) to ascertain the level of awareness of financial institutions outside the US. The survey was sent to the IIF, who then forwarded it electronically under a covering note to around eighty members of their economics advisory group that covers a wide range of financial institutions and investors from around the world. As noted, the response rate was too low to provide any meaningful data.

Finally, the survey was sent to two institutions (Oxford Analytica and the Estandardsforum) known to be active users of fiscal ROSCs as inputs to their own assessments of fiscal transparency; and also to the Economist Intelligence Unit.

APPENDIX II

II. Summary of Private Sector Survey: Quantitative Answers

Table 1 is based on 29 different responses from three categories of respondents. The overall response rate was 52 percent – although the response rates to some of the individual question were much lower. The first category includes rating agencies, the second are US-based international Banks and Investment Banks who were sent the surveys directly. The third, “Others,” includes Oxford Analytica, eStandardsforum and EIU. Table 1 shows response rates and mean responses for all the quantitative questions in the survey by each group and all respondents.

Table A1.Summary of Quantitative Questions in the Survey of Financial Institutions
ALL (29 of 56)1/RAT (21 of 42)INV/BNK(4 of 10)IIF(3 of 80)OTH (4 of 4)
RangePercent who

answered
Mean ResponsePercent for each response category 1/Percent who

answered
Mean ResponsePercent who

answered
Mean ResponsePercent who

answered
Mean ResponsePercent who

answered
Mean Response
012345
A. Familiarity with Fiscal Transparency
We you aware of the IMF Code on Fiscal Transparency?0=No, 1=Yes1000.8812881000.861001.001000.671001.00
Were you aware of fiscal ROSC reports?0=No, 1=Yes1000.8812881000.861001.001001.001001.00
Have you read part of a fiscal ROSC?0=No, 1=Yes1000.6832681000.711000.501000.671001.00
B. Use of Fiscal ROSCs
Better understanding of country’s fiscal position and risk

after reading fiscal ROSC
0=No, 1=Yes640.94694670.93501.00671.001001.00
Used info from fiscal ROSC as direct input0=No, 1=Yes680.534753710.53500.50670.501000.75
Followed up info from a fiscal ROSC or cited0=No, 1=Yes 0680.594159710.60500.50670.001000.50
Cited something in a fiscal ROSC to supervisory body or

others in org
0=No, 1=Yes680.594159710.60500.50670.501000.75
Significant impact on assessment1=No impact

5=Major

Impact
682.2918531866712.33502.00672.501004.50
Over past year, has your use of fiscal ROSCs:1=increased, 2=

same, 3=

decreased
681.7135596711.67502.00671.501002.00
C. Strengths and Weaknesses of Published Fiscal ROSCs
How important are the following features to you with

respect to fiscal ROSC?
1=Not important at all, 5=Very important
Wide Coverage across countries684.06012182447713.93505.00675.001004.75
Ease of accessibility684.4100123553714.47504.00675.001004.50
Consistency and comparability across countries684.3501261865714.27505.00674.001004.75
Source of information on fiscal management institutions

and practices in each country
684.410662959714.40504.50674.501004.25
Produced by an official international organization643.8806254425673.93503.50674.001003.75
Inclusion of IMF staff recommendations on reforms

needed
684.2400125335714.27504.00673.001004.00
Please indicate whether you agree or disagree that fiscal

ROSCs:
1=Strongly disagree, 5=Strongly Agree
Are too long682.9461853240712.93503.00674.001002.50
Lack candor or frankness682.59123535180712.67502.00674.001003.50
Are updated too infrequently682.8863529246713.00502.00673.001004.50
Should include more quantitative data683.88024121847713.93503.50675.001003.25
Should include a quantitative rating of fiscal transparency

in each country
683.8866242441713.80504.50673.501003.50

The responses from IIF and OTH groups were excluded from the calculation of response rates (percent answered) and mean responses for the ALL columns. The IIF group was excluded due to a very small response rate. The OTH group was excluded because they are active users of fiscal ROSCs as inputs to their own assessments of fiscal transparency.

The responses from IIF and OTH groups were excluded from the calculation of response rates (percent answered) and mean responses for the ALL columns. The IIF group was excluded due to a very small response rate. The OTH group was excluded because they are active users of fiscal ROSCs as inputs to their own assessments of fiscal transparency.

Table A1.(Continued). Summary of Quantitative Questions in the Survey of Private Sector
ALL (29 of 56) 2/RAT (21 of 42)INV/BNK (4 of 10)OTH (4 of 4)
RangePercent who

answered
Mean ResponsePercent for each response category 1/Percent who

answered
Mean ResponsePercent who

answered
Mean ResponsePercent who

answered
Mean Response
012345
D. Sources and Quality of Fiscal Data
How adequate to your needs do you consider the published

fiscal data to be?
1=very deficient, 5=

meeting all your needs
1003.10014622401003.101003.251003.00
Has the fiscal data published by the governments you

monitor improved in the last one to two years?
1=no improvement, 5=

major improvement
932.9343037300952.801003.25753.33
Assuming the information was generally available, please

rate the importance to you of the following types of fiscal

data and information
1=Not Important at all, 5=Very important
Public debt net of public financial assets1004.28331714621004.381004.751003.25
Civil service pension liabilities1003.970101738341003.951004.501003.50
Future social security pension obligations1003.93032841281003.901004.251003.75
Guarantees and other contingent liabilities1004.5900334621004.711004.501004.00
Off-budget expenditures1004.6600721721004.811004.251004.25
Tax expenditures1003.383173428171003.291003.251004.00
Non-financial performance info1003.00331342471002.951003.501002.75
Sensitivity of budget to changes in macro assumptions1004.34031721591004.521004.251003.50
Presentation of the govt, medium term fiscal outlook and

broad macro policy intentions
1004.45001034551004.521004.001004.50
Degree to which there are opportunities for legislative and

broader public debate and

input into fiscal policy and annual budget process
973.5747295011953.501003.751003.75
QFA by public and private financial inst1004.3807728591004.431004.501004.00
QFA by NFPE1004.14072124481004.141004.501003.75
In terms of the coverage of fiscal data, which measure do

you pay most attention to in practice?
Reported central government budget deficit0=No, 1=Yes970.079371000.101000.00750.00
Total central government deficit0=No, 1=Yes970.1189111000.141000.00750.00
General government deficit0=No, 1=Yes970.6832681000.901000.00750.00
Public sector deficit0=No, 1=Yes970.4357431000.291001.00750.67
Do you make adjustments to reported data on fiscal balance

and public debt.
0=No, 1=Yes930.8515851001.001000.50500.00
Do you consider that a more widespread use of accrual

basis reporting by government would improve your ability

to complete your investment, credit rating, or other financial

assesment?
1=No impact, 5=Major

impact
973.2914112532181003.381003.00753.00

The calculation of response rates (percent answered) and mean responses for the ALL category excludes responses from IIF. The IIF group was dropped from the overall analysis due to a very small response rate.

The calculation of response rates (percent answered) and mean responses for the ALL category excludes responses from IIF. The IIF group was dropped from the overall analysis due to a very small response rate.

Table A1(Concluded). Summary of Quantitative Questions in the Survey of Private Sector
ALL (29 of 56)1/RAT (21 of 42)INV/BNK(4of 10)OTH(4 of 4)
RangePercent who

answered
Mean ResponsePercent for each response category 1/Percent who

answered
Mean ResponsePercent who

answered
Mean ResponsePercent who

answered
Mean Response
012345
E. Fiscal Transparency Assessments by Private Sector Organizations
Aware of private sector assessments of fiscal

transparency
0=No, 1=Yes1000.4456441000.331001.001000.75
Have you ever read a private sector assessment?0=No, 1=Yes960.4258421000.38750.671000.75
Private sector assessment are1=superior,2=very

similar in

quality,3=inferior
282.4329071193.00751.67252.00
1=necessary,

2=unnecessary
241.001000141.00751.00751.00
1=duplicative,

2=complementary
361.891189292.00751.67752.00
Will you make more use of fiscal transparenc

material produced by IMF and others?
0=No, 1=Yes920.91991900.891001.00750.67
Short summary?0=No, 1=Yes960.752575950.751000.75751.00

The responses from IIF and OTH groups were excluded from the calculation of response rates (percent answered) and mean responses for the ALL columns. The IIF group was excluded due to a very small response rate. The OTH group was excluded because they are active users of fiscal ROSCs as inputs to their own assessments of fiscal transparency.

The responses from IIF and OTH groups were excluded from the calculation of response rates (percent answered) and mean responses for the ALL columns. The IIF group was excluded due to a very small response rate. The OTH group was excluded because they are active users of fiscal ROSCs as inputs to their own assessments of fiscal transparency.

A. Familiarity with Fiscal Transparency

This section was designed to assess the awareness of fiscal ROSCs in the private sector. As expected most of the respondents were aware of the IMF Code on Fiscal Transparency and ROSC reports with 88 percent of all respondents45 saying they were aware of the code on fiscal transparency. Similarly, 88 percent of the respondents were aware of the ROSC reports and 68 percent had read part of a ROSC.

B. Use of Fiscal ROSCs

The questions in this section attempted to gauge the extent of use of fiscal ROSCs amongst those who had read a fiscal ROSC. A significant portion of the respondents who answered (94 percent) felt that they had a better understanding of a country’s fiscal position and risk after reading a ROSC. On the use of information from fiscal ROSCs as a direct input to their assessment, the mean response was around 0.5, indicating that about half had used information from ROSCs directly. On a scale of 1 (signifying no impact) to 5 (being major impact), the average response was 2.3 for the impact of fiscal ROSCs on assessment, indicating some impact of fiscal ROSCs. Six respondents said a fiscal ROSC had a major impact (5) on the assessment, and six others responded with a 4.

C. Strengths and Weaknesses of Published Fiscal ROSCs

This section asked questions about the strengths and weaknesses of ROSC and was answered by only those who had read a ROSC before. The first part asked the respondents to choose from 1 to 5, 1 being not important at all to 5 being very important, regarding strengths of fiscal ROSCs. The majority (defined as more than 50 percent of the respondents choosing that category) of the respondents felt that very important aspects of fiscal ROSCs included ease of accessibility, uniformity across countries, and source of information on fiscal institutions and practices. A significant proportion also agreed that it was important that ROSCs included IMF staff recommendations.

In the next section the respondents were asked to choose from 1 to 5 (1 being strongly agree and 5 being strongly disagree) regarding some perceived weaknesses of fiscal ROSCs. A significant proportion of the respondents agreed that fiscal ROSCs should include more quantitative data (47 percent saying strongly agree) and that fiscal ROSCs should include a rating of fiscal transparency (41 percent saying strongly agree). On the question of whether the fiscal ROSCs were too long or if they lacked candor, most people chose in the range of 2 – 3 indicating somewhat neutral response.

D. Sources and Quality of Fiscal Data46

This section of the survey asked questions about the adequacy of fiscal data currently available and the kinds of data that respondents would find useful. On a range of 1 through 5 (1 being very deficient to 5 being meeting all needs), the average response was 3.1 indicating most respondents found the available data to be adequate. Most respondents reported some improvement in fiscal data in the last 1-2 years, with an average response of 2.9 on a scale of 1 to 5 (1 being no improvement and 5 being major improvement).

The next set of questions was designed to ascertain the kinds of fiscal data that the private sector would find useful. They were asked to rate different kinds of fiscal data from 1 (indicating not important at all) to 5 (indicating very important). Data for which the majority of respondents (defined as greater than 50 percent) indicated 5 were: public debt net of public financial assets; guarantees and other contingent liabilities; off-budget expenditures; sensitivity of the budget to macroeconomic risks; the medium term outlook and broad macroeconomic policy; and quasi-fiscal activities by public financial institutions and public enterprises. Amongst these, most people found off-budget expenditures to be very important, with 72 percent of the respondents choosing 5. Least interest was shown in data on non-financial performance information with an average response of 3.0.

Most people indicated that in practice they pay most attention to the general government deficit (followed by the public sector deficit). Over 80 percent of the respondents stated that they make some adjustments to reported data on the fiscal balance and public debt. The survey indicated that widespread use of accrual basis reporting will have somewhat of an impact on private sector assessments with an average response of 3.29 on a scale from 1 (no impact) to 5 (major impact).

E. Fiscal Transparency Assessments by Private Sector Organizations

Awareness of private sector transparency assessments was low with less than half the respondents (44 percent) saying they were aware of them. A small proportion of the respondents chose to answer questions comparing IMF fiscal ROSCs and private sector assessments. Of those who answered, a majority indicated that the private sector assessments were inferior, complementary, and necessary.

Summary of Responses from the “Others” Category

This group included responses from Oxford Analytica, Estandardsforum, and EIU. The first two of these institutions are fundamentally different from the other respondents to the survey because they use fiscal ROSCs as an input to their own assessments and ratings of fiscal transparency. (EIU was included to see they if they had considered doing the same thing). All respondents in this category were aware of the fiscal ROSCs and the Code. As expected, Oxford Analytica and Estandardsforum said that they had cited something in a fiscal ROSC and also that fiscal ROSCs had a significant impact on their assessments. They all agreed with the major strengths of fiscal ROSCs except they felt that it was not important that they were produced by an official international organization. Oxford Analytica and Estandardsforum felt that private sector assessments were very similar in quality to IMF ROSCs, necessary, and complementary.

Q3 attempted to gauge the extent of use of fiscal ROSCs amongst those who had read a fiscal ROSC. The last section of Q3 asked respondents to list countries on which their answers were based and to provide any additional comments.

Most of the respondents named emerging market economies such as the Philippines, Turkey, Brazil, Tunisia, and the Czech Republic. Some of the reasons for turning to a fiscal ROSC included: it is a source of reference particularly when starting on analysis of a country; to better understand the fiscal data and to fill some gaps in data. Many people pointed out better understanding of the extent of fiscal control, off-budget activities, and quasi-fiscal activities. One or two respondents pointed out that it is difficult to use fiscal ROSCs consistently because they are not uniform across countries, and because of resource constraints.

In the last section of Q5, respondents were asked to specify some other features of fiscal ROSCs that they consider to be important, in addition to the features already referred to in Q% (such as wide coverage or ease of accessibility).

Most respondents chose not to respond, but many who did indicated ROSCs should provide an indication of the true size of the fiscal deficit, and a true picture of fiscal credibility. One was the role of fiscal ROSCs in providing information on what may be missing from government estimates such as SOE borrowing or QFAs. Some respondents felt that Fiscal ROSCs should give an assessment of the quality of information, the track record of accounting and reporting practices, the impact of fiscal deficiencies on macro performance, and past management on NFPEs and QFAs. One respondent said there needs to be a closer linkage between the findings of ROSCs and the design of fiscal conditionality in IMF programs.

Q6. Comments: Please provide any additional comments on the quality of any fiscal ROSCs you have read.

A few users suggested that ROSCs lacked comparability and consistency across countries. Two respondents suggested that in order to facilitate comparisons, ROSCs should be organized more along the lines of the fiscal transparency code rather than being organized according to important issues in the country. Another suggested that ROSCs should include a ranking based on countries’ compliance with the code. One user felt it would be useful to have actual data in ROSCs, while another commented that it is important for ROSCs to be published quickly instead of letting the countries delay publication until they have addressed all the criticisms.

Q7. From what main sources do you obtain fiscal information on the country you monitor? Please list in the order of importance:

Most respondents mentioned the IMF and World Bank as major sources along with some other international organizations like ADB. Country included reports by Ministry of Finance, Statistical Agencies and Central Banks. Some respondents also use information from local economists, private sector assessments, rating agencies, IIF, and EIU.

Q9. What do you consider to be the key deficiencies of the fiscal data that you use?

Many respondents mentioned lack of comparability and consistency both historically and across countries as a key deficiency of available fiscal data. The lack of comparability is due to different accounting bases, different definitions of the government sector, and different treatment of various one-off transactions. Another common concern was lack of timeliness and frequency of the data. There was some concern about comprehensiveness of the data, especially the lack of data beyond the central government, and gaps with respect to QFAs, off-budget activities, and contingent liabilities.

Q11. If there has been an improvement in fiscal data, please give specific examples of how it has improved for particular countries.

Many respondents mentioned that accessibility of data has improved due to use of information technology, particularly in Taiwan, Sri Lanka, and Philippines. Other countries that were mentioned where fiscal data had improved were EU accession countries, Cyprus, Peru, Bahrain, Indonesia, Brazil, Egypt, South Africa, and Thailand. The two primary reasons mentioned for improvements mentioned were “reaction to crisis” (Brazil and many Asian countries) and “adopting EU practices in-line for accession” (Eastern Europe).

Q13. In terms of coverage of fiscal data, which measure of the fiscal balance do you pay most attention to in practice? Please provide explanation as appropriate.

Most respondents said they pay most attention to general government balance because it captures many of the aspects missed by the central government balance. Two countries mentioned where central and general government balance differed significantly were India and Argentina. It was also felt that it is more appropriate to compare general government deficit across countries since it is more likely to be consistent and comprehensive. However, a number indicated that the public sector balance was also important, while some said in practice they used the central government balance due to lack of data beyond central government.

Q14. Are there any adjustments that you typically make to reported data on the fiscal balance or public debt, to produce numbers that you regard as better measures? If yes, please indicate the adjustments typically made.

The most common adjustment was subtraction of privatization proceeds from revenues and treated as financing items. Other adjustments mentioned were, adding sub-national government data, and taking government guarantees into account in debt calculations.

Q15. Do you consider that a more widespread use of accrual basis reporting by governments would improve your ability to undertake you investment, credit rating, or other financial assessment? Please give reasons for you response.

Many people were of the view that although accrual accounting is superior in principle, due to practical considerations a switch to accrual accounting will not have a major impact on their assessments. One of the reasons mentioned was that many governments will continue to report on a cash basis making it difficult to compare across countries. Resources may be better spent improving the comprehensiveness and timeliness of cash accounting measures. Countries reporting on an accrual basis need to provide a full reconciliation to cash-basis numbers.

Q16. In this question the respondents were asked to make comments on fiscal transparency assessments being produced by the private sector, in comparison to fiscal ROSCs.

Respondents who answered this question felt that it was useful to have private sector assessments to complement IMF ROSCs, as they provide a diversity of views and approaches, are more concise, and provide a rating of transparency.

Q17. Please list any other assessments of fiscal or budget transparency that you are aware of produced by the private sector, other official organizations (e.g. the World Bank, the OECD) or by NGOs (e.g. the CBPP), indicating those that you found most useful.

The most frequently mentioned source was the OECD (a variety of publications) for developed countries. Many also mentioned World Bank publications, while some referred to the IMF, Estandards, and one referred to two civil society publications by the International Budget Project.

Q18. What are your views on the desirability, in principle, of publicly available assessments of fiscal transparency being produced by private (non-government) sector entities versus public (official) sector entities? Please briefly describe you views.

A number of respondents saw an advantage for private sector assessments in that they can be more clear and frank since they are more independent of governments, and there may be fewer political considerations. Many respondents felt that official and private sector assessments complement each other, and it is desirable to have both. Some of the reasons why official sources may have an advantage in producing these assessments were: it is not clear if the private sector can make money from ROSCs; the amount of resources needed to do quality reports may be prohibitively high for the private sector; and the official sector may have more access to information and expertise.

Q19. Do you consider you may in future make more use than in past of published material on fiscal transparency produced by IMF or others. If yes, please briefly indicate which material, and how you might use it.

Some respondents indicated they will read fiscal ROSCs (and other IMF reports) on a regular basis, especially as more ROSCs become available, and especially for countries where information is otherwise hard to obtain.

Q20. Finally do you have any suggestions on how the IMF’s fiscal transparency initiative could be made more useful to you. Please indicate areas of suggested improvement, ranking your suggestions in order of importance.

Many respondents felt that country coverage should be increased, along with more frequent updates. Many also suggested that fiscal ROSCs should include estimates of the true deficit and debt, and to indicate the scale of particular off-budget transactions rather than just describing them. A few people mentioned having a standardized score or a numerical rating included in the fiscal ROSC. It was suggested that fiscal ROSCs should be organized according to the code so it is easy to compare across countries, a summary table comparing compliance for each part could be included, as well as an annex showing compliance in a comparable group of economies.

APPENDIX III

III. Financial Institutions Interview Schedule1

Rating Agencies:

Standard and Poors: Marie Cavanaugh and Joydeep Mukherji, Directors, Sovereign Ratings. New York.

Moody’s Investor Services: Thomas Byrne and Mauro Leos, Vice Presidents, Sovereign Risk Unit. New York.

Fitch Ratings: Lionel Price, Chief Economist, and Richard Fox, Senior Director, Sovereign and International Public Finance. London.

New York-based Investment Banks:

Citibank/CitiGroup: Cheryl Rathbun, Vice President and Chief of Staff, Global Corporate and Investment Bank.

JP Morgan: Alexei Remizov and Leanna Bryerlee, Vice Presidents, Country Risk.

Bank of New York: Doug Chapman, Senior Vice President, and John Koch, Vice President.

Organizations Independently Assessing and Disseminating Information on Fiscal Transparency:

Oxford Analytica: Michael Bates, Director of Consultancy and Research, and Juan Garin, Deputy Director. Oxford.

Estandardsforum: George Vojta, Chairman, and Matt Zimmer, Principal, with Carl Adams, Executive Director, Financial Standards Foundation. New York.

Other:

Economist Intelligence Unit: Merli Baroudi, Director Country Risk Service, David Anthony, Director, Country Risk Alerts, Matthew Sherwood, Senior Editor/Economist, and Andre Astrow, Deputy Editorial Director. London.

Institute for International Finance: Yusuke Horiguchi, First Deputy Managing Director, and Sabine Miltner, Senior Advisor and Director, Multilateral Policy Department. Washington DC.

APPENDIX IV

IV. Civil Society Surveys

Two separate surveys were distributed:

  • One to the Center for Budget and Policy Priorities (CBPP) in Washington DC, and, via the CBPP and under a covering note from them, to 48 of their partner organizations in the International Budget Project (IBP) in 22 countries – referred to as the IBP survey; and

  • One to 58 development civil society organizations (CSOs) in developed and developing countries, from lists maintained by the IMF - referred to as the CSO survey.

The surveys were distributed by email, and respondents were asked to complete the surveys electronically and return them by email. Table 2 contains a summary of the quantitative results.

Table A2.Summary of Quantitative Questions in the Survey of Civil Society Organizations and Budget-Related NGOs
Budget related NGOs (5 of 48)Development CSO (10 of 58)
RangeNumber who answeredMean ResponseNumber who answeredMean Response
Familiarity with Budget and Fiscal Transparency
Were you already aware of the IMF Code of Good Practices on Fiscal Transparency?0=No, 1=Yes50.80100.20
Were you already aware of Fiscal Transparency country reports (ROSCs) published by

the IMF?
0=No, 1=Yes50.80100.20
Have you used or drawn on information in the Code or fiscal ROSC in any way in your

analysis or advocacy work?
0=No, 1=Yes50.8030.33
Have you read some part of a fiscal ROSC?0=No, 1=Yes50.2030.33
Sources and Quality of Budget and Expenditure Data
How adequate to your needs do you consider the published budget and expenditure data

that is available to you?
1=very defficient, 5=

meeting all your

requirements
52.00
Has the budget and expenditure data published by the government(s) you monitor

improved in the last one or two years?
1=no change, 5=

major improvement
41.75
Budget Transparency Assessments by NGOs
Were you aware of NGO studies on budget transparency?0=No, 1=Yes41.00100.40
Have you read a NGO study of budget transparency?0=No, 1=Yes40.75100.30
Future Directions
Do you anticipate that you may make more use in future of material on budget

transparency published by IMF than you do currently?
0=No, 1=Yes41.00101.00
The Fund could perhaps more actively publicize and disseminate its fiscal transparency

intiatives and fiscal ROSCs. Please rate how useful you consider the following intitiatives

would be:
1=least useful, 5=most useful
Presentations to civil society groups on Fiscal Tranparency54.40104.00
Open-forun discussion with civil society on Fiscal Transprency54.20104.00
Issuing a Public Information Notice periodically indicating fical ROSCs that have been

posted on the IMF’s web sit in the previous period.
53.80105.00
Notification each time a new fiscal ROSC is posted on the Fiscal transparency website54.20
Issuing a press notice in-country when a fiscal ROSC is published on that country.54.40105.00

A. Budget Related CSOs Survey

The IBP survey was intended to assess the level of awareness amongst applied budget analysis CSOs of the Fiscal Transparency Code and Manual; the extent of awareness and use of fiscal ROSCs; views on the adequacy of fiscal data; and views on how the Fund should more actively disseminate information on its fiscal transparency initiatives.

Five responses were received to the IBP survey resulting in a response rate of around 10 percent. Many (4 out of 5) of the CSOs were aware of the fiscal transparency code and report on observation of standards and codes (ROSCs) but very few (1 out of 5) had actually read part of a ROSC. The respondents reported that there has been little improvement in budget and expenditure data published by the government they analyze, and that it is insufficient for their needs. As expected, of those who answered all said they were aware of the CSO studies of on budget transparency and most had also read a CSO study. Asked about how the IMF could publicize and disseminate its fiscal transparency initiative, most interest (4.4 average response out of 5) was shown in presentations to civil society groups, and issuance of in-country press notices.

B. Civil Society Organizations Survey

The CSO survey was intended to assess the level of awareness of the Code and fiscal ROSCs, and of CSO budget transparency assessments amongst the broader international CSO community; to obtain views on how the Fund should more actively disseminate information on its fiscal transparency initiatives; and to see if respondents anticipated they might in future make more use of IMF information on fiscal transparency than they currently do.

Ten responses were received to the CSO survey, a response rate of around 17 percent. Six of the respondents were from developing country CSOs, and four were U.S.-based. A small proportion (2 out of 10) of the respondents were aware of the fiscal code and fiscal ROSCs. About one third of the respondents were aware of CSO studies of fiscal ROSCs and had actually seen a CSO study. All respondents indicated that they anticipate in future making more use of fiscal ROSCs than they currently do. Asked about how the IMF could better publicize and disseminate its fiscal transparency initiative, most interest (average response of 5 out of 5) was shown in issuance of in-country press notices and periodic public information notices indicating all the ROSCs posted on the IMF’s website in the previous period. Responses were also high (average response of 4) for presentation on fiscal transparency to civil society groups and open-forum discussions with civil society.

References

    AllumPeter F. and MehmetAgca2001Economic Data Dissemination: What Influences Country Performance on Frequency and Timeliness?IMF Working Paper 01/173 (Washington: International Monetary Fund).

    BhatiaAshok Vir2002Sovereign Credit Ratings Methodology: An EvaluationIMF Working Paper 02/170 (Washington: International Monetary Fund).

    ChambersJohn2002The Importance of the IMF’s Work on Standards and Codes,” speech delivered at the “Standards and Transparencyseminar at the IMF/World Bank 2002 Annual MeetingsWashingtonSeptember.

    Code of Good Practices on Fiscal Transparency: Declaration on Principles1998IMF SurveyApril27 pp. 12224.

    CreditLyonnais2002Unravelling Public Finances: Gordian Debt” (Hong Kong: Credit Lyonnais Securities Asia).

    European Central Bank2002Recent Developments in International Co-OperationECB Monthly Bulletin (February) pp. 5365.

    Financial Stability Forum2000Report of the Follow-Up Group on Incentives to Foster Implementation of Standardspresented at the fourth FSF meetingBaselSeptember. Available via the Internet: http://www.fsforum.org/publications/Report_Incentives01.pdf.

    Financial Stability Forum2001Final Report of the Follow-Up Group on Incentives to Foster Implementation of Standardspresented at the sixth FSF meetingLondonSeptember. Available via the Internet: http://www.fsforum.org/publications/Final Incentives01.pdf

    FölscherAltaed.2002Budget Transparency and Participation: Five African Case Studies (Cape Town: Institute for Democracy in South Africa (IDASA))

    FölscherAltaWarrenKrafchik and IsaacShapiro1999Transparency and Participation in the Budget Process: South Africa: A Country Report (Cape Town: Institute for Democracy in South Africa (IDASA)).

    GelosR. Gaston and Shang-JinWei2002Transparency and International Investor BehaviorIMF Working Paper 02/174 (Washington: International Monetary Fund).

    GuerreroJuan Pablo and HelenHofbauer2001Index of Budget Transparency in Five Latin American Countries: Argentina, Brazil, Chile, Mexico and PeruInternational Budget Project (Washington: Center on Budget and Policy Priorities). Available via the Internet: http://www.internationalbudget.org/resources/1-10-02ibp-pr.htm.

    HemmingRichard and MichaelKell2001Promoting Fiscal Responsibility: Transparency, Rules and Independent Fiscal AuthoritiesFiscal Rules: Papers Presented at the Bank of Italy Workshop Held in Perugia 1-3 February 2001 (Rome: Banca d’Italia).

    HemmingRichard and MurrayPetrie2000A Framework for Assessing Fiscal VulnerabilityIMF Working Paper 00/52 (Washington: International Monetary Fund).

    Institute of International Finance1999Report of the Working Group on Transparency in Emerging Markets Finance (Washington: IIF).

    Institute of International Finance2002Action Plan of the IIF Special Committee on Crisis Prevention and Resolution in Emerging Markets” (Washington).

    International Monetary Fund1999Fiscal Transparency: An Experimental Report” in Greece: Selected Issues IMF Staff Country Report No. 99/138 (Washington).

    International Monetary Fund2001aManual on Fiscal Transparency (Washington). Available via the Internet: http://www.imf.org/external/np/fad/trans/manual/index.htm.

    International Monetary Fund2001bAssessing the Implementation of Standards—An IMF Review of Experience and Next StepsPublic Information Notice (PIN) No. 01/17 (Washington). Available via the Internet: http://www.imf.org/external/np/sec/pn/2001/pn0117.htm.

    International Monetary Fund2002Quarterly Report on the Assessments of Standards and Codes—June 2002” (Washington). Available via the Internet: http://www.imf.org/external/pubs/ft/stand/q/2002/eng/062102.htm.

    International Monetary Fund2003aInternational Standards—Background Paper on Strengthening Surveillance, Domestic Institutions, and International Markets” (Washington). Available via the Internet: http://www.imf.org/external/np/pdr/sac/2003/030503s1.pdf.

    International Monetary Fund2003bAssessing and Promoting Fiscal Transparency: A Report on Progress” (Washington). Available via the Internet: http://www.imf.org/external/np/pdr/sac/2003/030503s2.pdf.

    International Monetary Fund and World Bank2003aBank/Fund Collaboration on Public Expenditure Issues” (Washington). Available via the Internet: http://www.imf.org/external/np/fad/pubexpen/2003/021403.pdf.

    International Monetary Fund2003bInternational Standards: Strengthening Surveillance, Domestic Institutions, and International Markets” (Washington). Available via the Internet: http://www.imf.org/external/np/pdr/sac/2003/030503.pdf.

    JeunemaîtreAlain1997Financial Markets Regulation: A Practitioner’s Perspective (Houndmills, Basingstoke, Hampshire: Macmillan Press).

    KopitsGeorge and JonCraig1998Transparency in Government OperationsIMF Occasional Paper No. 158 (Washington: International Monetary Fund).

    LazereEd2002Improving the Transparency of the DC Budget” (Washington: District of Columbia Fiscal Policy Institute).

    LevichRichard M.GiovanniMajnoni and Carmen M.Reinharteds.2002Ratings Rating Agencies and the Global Financial System (Boston: Kluwer Academic Publishers).

    MisiagWojciech and AdamNiedzielski2001Openness and Transparency of Public Finance in Poland in the Light of International Monetary Fund StandardsIBnGR Working Paper No. 29/2001 (Warsaw: Gdansk Institute for Market Economics). Available via the Internet: http://www.internationalbudget.org/resources/library/Poland.pdf.

    MathiesonDonald and GarySchinasi1999International Capital Markets: Developments, Prospects, and Key Policy IssuesWorld Economic and Financial Surveys (Washington: International Monetary Fund).

    MosleyLayna2003Global Capital and National Governments (Cambridge: University of Cambridge Press).

    PetrieMurray1999The IMF Fiscal Transparency Code: A Potentially Powerful New Anti-Corruption Toolpaper presented at the 9th International Anti-Corruption ConferenceDurbanOctober. Available via the Internet: http://www.transparency.org/iacc/9th_iacc/papers/day4/ws2/dnld/d4ws2_mpetrieimf. pdf.

    PopeJeremyed.2000National Integrity Systems: The TI Source Book 2000 (Berlin: Transparency International).

    PriceLionel2002Standards and Codes—Their Impact on Sovereign RatingsFitchRatings Special Report10July 2002. Available via the Internet: http://216.239.39.104/search?q=cache:4chLA7oW8S8J:www.fitchratings.es/Informes/Standards%2520%26%2520Codes%2520-%2520Their%2520Impact%2520on%2520Sovereign%2520Ratings.pdf+Standards++ and+Codes--their+impact+on+Sovereign+Ratings&hl=en&ie=UTF-8.

    SchneiderBenu2002Issues in Implementing Standards and Codespaper presented at the Overseas Development Institute conference “International Standards and Codes: The Developing Country Perspective,”LondonJune. Available via the Internet: http://www.g24.org/schnetgm.pdf.

    Standard and Poor’s2002aSovereign Credit Ratings: A Primer. Available via the Internet: www.standardandpoors.com.

    Standard and Poor’s2002bGlobal Financial System Stress Appendix 5: Financial System Stress and Sovereign Credit Risk. Available via the Internet: www.standardandpoors.com.

    SteinErnestoErnestoTalvi and AlejandroGrisanti1998Institutional Arrangements and Fiscal Performance: The Latin American ExperienceIADB Working Paper 367Washington: Inter-American Development Bank). Available via the Internet: http://www.iadb.org/res/publications/pubfiles/pubWP-367.pdf.

The author thanks Farhan Hameed for assistance in the analysis and writeup of the survey results; Bill Allan and Taryn Parry for their input into the project and their comments on previous drafts; and Lynn Aylward, Anton op de Beke, Jim Brumby, Joel Friedman, George Kopits, Warren Krafchik, Lionel Price, Kristin Roesser, Jan Aart Scholte, Matt Zimmer, and participants in the “Discussion of Fiscal Transparency” session held during the IMF and World Bank’s 2003 Spring Meetings for helpful comments. Mr. Petrie is an independent consultant and a member of the panel of fiscal experts of the IMF. He is also an Executive Officer for Transparency International New Zealand, and has contributed to the budget transparency work of the Center for Budget Policies and Priorities. Comments on this draft should be sent to Mr. Petrie via e-mail.

More recently, a twelfth standard and code was recognized by the IMF’s Executive Board.

Previous surveys of financial market participants on their use of ROSCs have covered all the standards and codes as a group (see FSF, 2001, and IMF, 2003c). The standards are very diverse, with quite distinct communities of interest. It is difficult to clearly discern views of fiscal ROSCs from these exercises. The surveys have also been confined to the financial and official sectors.

The Code was amended in 2000, in light of initial experience in applying it. The main change related to strengthening the provisions relating to data quality. In recognition that one size does not fit all, the Fiscal Transparency Manual also identifies two other benchmarks, a set of basic requirements, and a set of best practices. (See IMF, 2001, overview.) The OECD has also promulgated a set of Best Practices for Budget Transparency. While broadly similar to the IMF Best Practices, they focus only on central government, and on the budget sector rather than all fiscal activities. They are not intended to constitute a formal “standard,” and the OECD is not assessing the extent to which countries adhere to them.

The IMF Board decided, at its 2001 review of experience in assessing standards, that in cases where countries’ observance of standards is deemed poor, or where the information available to the Board was insufficient, the Board would encourage countries to participate in a ROSC (IMF, 2001b, p. 8). Since then the Board has explicitly encouraged three countries (Ireland, Israel, and Mexico) to volunteer for a fiscal ROSC, and has also suggested that fiscal ROSCs for the members of the Eastern Caribbean Currency Union would be a helpful means of facilitating peer review of members’ achievement of fiscal benchmarks.

See IMF 2003b for a progress report on the implementation of the fiscal transparency initiative. See also IMF and World Bank 2003b and IMF 2003a for an assessment of the extent to which all ROSCs have influenced surveillance and TA.

See Bhatia, 2002, p. 52.

The sovereign ceiling refers to the fact that it is generally very difficult for any entity in a country to be rated higher than the government’s foreign currency rating, due to the government’s powers of taxation and foreign exchange control. The sovereign ceiling no longer seems to be regarded as a binding constraint in all instances, but it nevertheless remains a key factor. See Standard and Poors 2002a. See also Bhatia, 2002, for a discussion of the key role the sovereign ratings assigned by the three major rating agencies play in international capital markets.

See for instance Standard and Poors 2002a, Standard and Poors 2002b. See also Bhatia, 2002, who states that Standard and Poors responded to the Asian crisis by adding a third fiscal score to its rating categories, aimed at quantifying off-budget and contingent liabilities; and by splitting a combined category for external debt into two separate categories for public and private sector external debt.

There is generally less data available on emerging markets, and the data are less reliable. Investors evaluating sovereign risk in emerging markets also look for a wider range of information than they do with respect to developed country markets, in part because ability to repay and even willingness to repay are more important issues than they are with respect to developed countries.

A 1999 survey of rating agencies by the IMF found that on average each sovereign ratings analyst was responsible for rating seven countries (see IMF, 1999, p. 197). The rating agencies reportedly increased the resources devoted to sovereign ratings following the Asian crisis, and Bhatia (2002) states that each Standard and Poors and Fitch analyst is responsible for between four and five countries.

The Estandardsforum is a partnership between Oxford Analytica, the Wharton Financial Institutions Center, the Reinventing Bretton Woods Committee, and Mr. George Vojta (who is the Chairman).

The Financial Stability Forum (FSF) observed with respect to such gap-bridging initiatives: “The Group considers, overall, that the official sector should certainly not impede these initiatives, given their potential value, but neither should it imply its endorsement of the information they provide.” (FSF, 2001).

Oxford Analytica is an international consulting firm drawing on over 1,000 senior faculty members at Oxford and other major universities and research institutions around the world.

Other sources of assessment may be used if they meet certain criteria, such as being authoritative, objective and comprehensive. In one case, the Estandardsforum cites one of the CBPP studies (see Section IV) as the source of one of its country fiscal transparency ratings.

The Estandardsforum itself assigns a numeric rating to each of the six categories on a scale of 0-10 (no assessment available is given the zero score, while no compliance is given a 1, on the basis that no information is the worst possible situation facing international investors). They then sum these scores across all the 12 standards and codes to provide a snapshot and league tables of countries’ overall compliance with standards and codes.

A recent study by Gelos and Wei (2002) found that emerging market equity funds tend to allocate less money to less transparent countries, controlling for other factors. Their measure of country transparency included three components, one of which was ratings derived from Oxford Analytica reports of fiscal and monetary policy transparency against the respective IMF Codes. Gelos and Wei also found that herding among funds is somewhat less prevalent in more transparent countries; and they found support for the view that, during the Asian and Russian crises, international investors tended to flee less transparent countries. In contrast, they found that mutual funds react less strongly to macroeconomic news in less transparent countries, and more strongly to the same news in more transparent countries.

See CLSA, 2002.

See FSF, 2001, and IMF, 2003c.

See also Price, 2002, and Chambers, 2002, for rating agency comments on the relevance of ROSCs, including fiscal ROSCs.

Of the twenty survey responses, only four were from Moody’s. Given that Moody’s has not formally incorporated use of fiscal ROSCs into its operating procedures, this may mean there is some upward bias in the survey results on the level of awareness and use of ROSCs.

While fiscal ROSCs are increasingly incorporating links to official government web sites where fiscal data are posted, they do not themselves contain quantitative data.

This would be a more relevant and useful disclosure than the current disclosure, which states that the Fund’s publication policy allows for the deletion of market-sensitive material. This has very little if any relevance to a fiscal ROSC.

As occurred with the one-off fiscal transparency assessment exercise completed by CLSA.

In a similar manner to governments meeting the costs of the credit assessment services of bond rating agencies.

While some of the CalPERS-financed studies are available on its web site, not all of them are in the public domain. The Estandardsforum information, and the CLSA study have been available only on a subscription basis - although the Estandardsforum data is now publicly available.

Concerns have been expressed during official outreach exercises about the objectivity (and accuracy) of the assessments and information provided by private sector firms with respect to various international standards (see FSF 2001, and IMF 1999).

The recent IMF Board discussion of the Standards and Codes initiative noted a growing demand for ROSCs, and a need for greater prioritization of new ROSCs and up-dates of existing ROSCs, given limitations on available resources. Priority for new ROSCs is to be given to countries where the returns are greatest in terms of financial stability or development impact. The Fund is also exploring greater external partnerships in producing ROSCs and ROSC updates. See IMF Public Information Notice No. 03/43, April 3, 2002.

These suggestions were made by the IIF with respect to SDDS compliance. See IIF, 1999.

See IMF, 1999, Hemming and Petrie, 2000, and the Fiscal Management Assessment Report on Turkey, SM/02/191, 6/20/02, discussed in IMF, 2003a, p. 18.

See for instance FSF, 2001, IIF, 1999 and Schneider, 2002. The FSF did note that lack of political will within governments and legislatures was proving to be a barrier to the further implementation of standards, and suggested efforts to directly reach the business community within these countries as a way of building a reform-minded constituency. The FSF did not refer, however, to the potential role of the broader civil society.

For general information on the International Budget Project and summaries of civil society budget transparency studies, see the IBP web site, www.internationalbudget.org/themes/BudTrans/transp.html

For the DC study, see Lazere, 2002. Information on both the Russian project and the DC study is available on the IBP web site.

See Pope, 2000, Chapter 23. See also Petrie, 1999.

See TI’s Press Release on the Global Corruption Report 2003, 22 January 2003.

The list of countries where there is both a fiscal ROSC and a civil society study omits Argentina, where there is only a very short, early experimental fiscal ROSC.

See the International Budget Project Newsletter, No. 13, January 2003. Available at www.internationalbudget.org/resources/newsletter13.htm

For example, IMF staff missions to the Czech Republic (a country with a high number of ROSCs) have regularly flagged standards and codes issues with the local press while the former IMF resident representative for the Czech Republic and Hungary gave presentations on standards at seminars. See IMF, 2003a.

Similar to the one published recently - see IMF, 2003d.

The overall response rates and mean responses were calculated excluding the 4 respondents from the “Others” category.

Calculations for this section included the responses from “Others” category.

The meetings in New York were attended by Mrs. Parry and Mr. Petrie, the meetings in London and Oxford were with Mr. Allan, and the meeting with the IIF was attended by Messrs. Allan, Bertuch-Samuels and Petrie, and Mrs. Parry.

Other Resources Citing This Publication