Why is transparency about public expenditure so elusive?

1. The claims of transparency

Claims made about transparency, and in the name of transparency, are pervasive in current public debate. Transparency is often presented as a public virtue, which it is discreditable or inadvisable to oppose. Usually, transparency is something that the advocate wishes to impose on someone else or some organization, often a public one. The presumed relationship is asymmetric, with transparency demanded on issues about which the demander would not expect to have to reciprocate. Making claims about transparency appears to be consistent with a very wide range of ideological views and substantive policy proposals. ‘Transparency’ slips easily off the tongue and pen, often in tandem with ‘openness’, without it being clear whether these represent different concepts or are synonyms used for emphasis. Transparency is much acclaimed but its effects are more ambiguous than is suggested by contemporary portrayal.

Justice Brandeis’ famous claim that ‘sunlight is the most powerful of all disinfectants’ (Freund, 1972) has great rhetorical power but does not advance the analysis very far. It does, however, prompt an important distinction as to whether the expectation of transparency as sunshine is to drive out corruption or whether the goals are greater efficiency and effectiveness from the use of public resources and/or enhanced legitimacy and accountability of public institutions. This raises the question as to whether the dominant political concern in a particular country is, on the one hand, about reducing corruption or, on the other, about increasing efficiency, effectiveness and legitimacy. The optimal configuration of transparency would be expected to differ between these two cases.

Although concerns for transparency – not necessarily using the word – have a long history (Hood, 2006), current claims for transparency target both ‘public’ and ‘private’ activities and personal and professional actions. Few political speeches or media interviews about public policy now omit an appeal to transparency as a non-negotiable consideration. Some of these appeals re-clothe enduring interests and ideas in fashionable garb, much like earlier appeals to ‘democracy’ or ‘accountability’. Some project transparency as the solution to governance problems, the imprecision of both concept and mechanisms giving it the appearance of panacea, making it difficult to counter.

Notwithstanding that part of the transparency phenomenon is political and media knockabout, something important has happened. Transparency is associated with developments in auditability and disclosure (Power, 1999). It is influenced by technological change (e.g. growth of the internet and surveillance technology) and by media developments (e.g. concentration and trivialization). It is linked to the changing nature of the state and particularly to the way in which a self-contained state, directly undertaking production and redistribution (positive state), has to varying degrees been replaced by a regulatory state (Majone, 1997). This is characterized by greater reliance on indirect means of achieving public policy objectives, thus more dependent on reliable information flows between organizations and less on hierarchical command-and-control within them. Transparency is often believed to address deficits in trust, credibility and legitimacy, although there are contrary views: for example, O’Neill (2006) views transparency as destructive of trust, particularly in professional expertise.

This expanding public profile of transparency is stimulating the development of a diverse academic literature, exemplified by the number of disciplines, topics and methodologies represented at the First Global Transparency conference, held at Rutgers University in May 2011. In terms of the role of transparency in public administration, influential voices have been Mitchell (1998), Hood (2001), Roberts (2006), Bovens (2007) and Piotrowski (2007). There is no space in this article to summarize such contributions, but indebtedness to these precedents is acknowledged.

The central argument of this article is that the way in which transparency mechanisms are structured will shape their impact on public policy – on effectiveness on equity and on democratic accountability. The article sheds light on the transparency of both process and substance in relation to public expenditure, the outflow side of the public budget. It uses the experience of the United Kingdom, which scores highly on international rankings of fiscal transparency, as a vehicle for highlighting generic issues, including the dangers of placing too much weight on measurable indicators in the absence of an explicit conceptualization of fiscal transparency.

The article is structured in the following way. This section has considered the claims made for transparency, which is often presented as an imperative. Section 2 provides a conceptualization of transparency which is useful in considering the specifics of transparency in a fiscal and public expenditure context. Section 3 introduces the role of transparency in the fiscal domain. Having established this background and the tools of analysis, attention turns in section 4 to transparency about public expenditure. Section 5 highlights why recourse to surrogates for public expenditure must be monitored. Finally, section 6 concludes the analysis of why public expenditure transparency is elusive, while articulating some general principles that should influence practical steps to increase transparency towards levels achieved in best-practice countries.

2. The conceptualization of transparency

Heald (2003) focused on a transparency trade-off between the ‘value of sunlight’ (e.g. flushing out incompetence and corruption) and the ‘danger of over-exposure’ (e.g. avoiding excessive politicization or dysfunctional surveillance). While this formulation provides a valuable starting point, it is incomplete because it does not distinguish varieties of transparency. In particular, it formulates the design problem in terms of ‘too much or too little transparency’ rather than ‘desirable and undesirable (mixes of) varieties of transparency’. Prat (2005) highlighted the damaging effects of the ‘wrong kind of transparency’.

As shown in Figure 1, Heald (2006a) conceptualized the ‘directions of transparency’ as upwards/downwards (the vertical dimension) and inwards/outwards (the horizontal dimension). The vertical dimension can be thought of in principal–agent terms. Transparency is a mechanism through which the principal can exercise surveillance over the actions of an agent. If the relationship is symmetric, the agent can also view the actions of the principal. Otherwise the vertical transparency relationship is asymmetric. OPEN IN VIEWER

The direction of the principal–agent relationship in the vertical dimension of transparency may be disputed. Heald (2006a) reflects the hierarchical nature of public organizations, so that, for example, employees are agents of managers who are in turn agents of directors who are in turn agents of ministers. Analogously, upwards transparency allows the rulers (ministers) to monitor what the ruled (as citizens and as employees) are doing; it is managerially focused. Downwards transparency is accountability focused, providing a vehicle for rulers to account to citizens for their stewardship. An alternative view, appealing to constitutional and democratic theory, would portray the citizens as the principal of the elected Legislature and ministers as agents. Adopting this alternative view would affect the presentation, but not the substance, of the vertical dimension.

The horizontal dimension is related to Jeremy Bentham’s panopticon (Hood, 2006) – actions inside the glass structure can be observed from outside. Bentham saw this as a means of efficient surveillance over the activities of prisoners. Again, the question of symmetry arises: is it possible from within the panopticon to view actions outside?

‘Fully symmetric’ transparency (Heald, 2006a: 27–29) occurs when all four directions are simultaneously present. This is a descriptive label, not a statement about optimality. The multiple directions of transparency clarify why claims about transparency can so easily be made, yet their substantive implications can remain ambiguous or contradictory.

While retaining in mind the four directions, the structure of transparency is characterized by four components in Figure 2 (Heald, 2006a: 30–37):

The distinction between event transparency and process transparency. In public policy, the former is concerned with the transformation of inputs into outputs into outcomes, on the assumption that these are measurable or can be proxied. Process transparency incorporates a distinction between procedural and operational aspects; it is much less intrusive on the internal space of an organization to have its procedures documented in the public domain than it is to have its operational practice constantly in public view.

OPEN IN VIEWER

Figure 2.

The structure of transparency

OPEN IN VIEWER

Figure 3.

Transparency in retrospect versus in real time

There can be differences of view about the relative weights to be placed on the structural components of transparency. In turn, these are contingent on the weights placed on those values that potentially compete with transparency and on the implicit model of organizational responses to transparency.

The argument is essentially about whether primacy should be given to transparency over other values. This reveals a distinction between transparency as an intrinsic and an instrumental value. Birkinshaw (2006) contended that transparency is a human right, while Heald (2006b) insisted that it should be regarded as an instrumental value. Public policy involves both synergies and trade-offs between valued concepts. Without claiming to be comprehensive, Heald (2006b: 67–68) identified seven other values which, in particular circumstances and habitats, transparency might support or compete with: effectiveness; trust; accountability; autonomy and control; confidentiality, privacy and anonymity; fairness; and legitimacy.

Resolving, or at least accommodating, these distinct sets of values enters Hood’s (1976) ‘management of dilemma’ territory. On this view, dilemmas cannot be resolved as such, but managed for a time and space by reference to the priorities of that time and space. This means that there are no absolutes; only relative responses to contingencies. It is therefore important to emphasize the moral ambiguity of transparency and not to accept the assumption underpinning much rhetoric that (greater) transparency is automatically beneficial. Much depends on what the transparency is about (event versus process) and contextual matters such as the power and legitimacy relationships between those who demand transparency and those who are asked to provide it.

Moreover, there is a difficult change of domain from discussions about transparency in relation to an individual to those about transparency in relation to public organizations and to individuals with public functions. An argument could be constructed that all directions and varieties of transparency might work against the ‘social functions of ignorance’ (Moore and Tumin, 1949). The ambiguity, fudge and suppressed memories that may lubricate difficult family relationships may also promote peace in zones of civil conflict as well as offering some protection against centralized political power. It is easy to construct examples from both personal and public domains where departures from transparency might be argued to be justified (‘better not to know’). However, it is difficult to determine how much counterweight to place on the social functions of ignorance while the potential for opportunist use against transparency claims is obvious.

3. Transparency in the fiscal domain

This section will highlight matters of broad principle, leaving the detail to sections 4 and 5. The governance of the fiscal domain is central to what it means to be a state, especially when that state has claims to be democratic. Fundamental to this is the distinction between resources belonging to the state and to the sovereign/ruler; this is understood in most industrial democracies but not in many other countries, with baleful effects on the fiscal domain.

Attitudes towards notions of accountability and transparency will depend on the explicit or implicit model that underpins one’s view of the fiscal activities of the state. Nowhere has this been so effectively evoked as in the 1998 Munich Lectures by James Buchanan and Richard Musgrave (Buchanan and Musgrave, 1999). While not unworldly about fiscal politics in a democracy, Musgrave viewed politicians and officials as attempting to maximize some conception of social welfare. In Buchanan’s bleaker view, politicians are rent-seeking vote maximizers and officials are rent-seeking budget maximizers, unleashing the force of Leviathan on citizen-taxpayers. On this public choice view, institutions and policy processes should be assessed in relation to their contribution to restraining the state. Instead of governments being endowed with broad-based taxes that minimize the excess burden of taxation, narrow tax bases entailing high excess burden are to be preferred, along with constitutional constraints on spending and deficits. The decision on whether to opt for fiscal transparency may be an instrumental decision, conditional on expected effects.

There are grounds for scepticism about high-level transparency indexes, including those in relation to public expenditure or budgeting based on self-reported country data to the OECD (Bastida and Benito, 2007). Indexes incorporate the presumptions of index creators, assume that transparency indicators are additive, and do not handle well the potential interactive effects. They are also vulnerable to marked divergences between nominal transparency (e.g. desirable structural features are in place) and effective transparency (e.g. hostile habitat or unsupportive cultural traditions exist). The International Monetary Fund (IMF) was commissioned in 2009 to rate G20 countries on the quality of their fiscal institutions, fiscal transparency being one dimension of quality. As a result of country objections to the results, this expert-rated study was never published.

The International Budget Partnership is a non-governmental organization that seeks to ‘contribute to reforms in how governments around the world manage public funds’. The 2010 Open Budget Index (International Budget Partnership, 2010: 9) scored the United Kingdom at 87 on a scale that runs from 0 to 100. This places it third of the 94 assessed countries, in an elite group of seven – behind South Africa and New Zealand but ahead of France, Norway, Sweden and the United States. This elevated position for the United Kingdom jars with evident weaknesses (Heald, 2003, and this article). Although it ticks many of the indicators, which inevitably relate to formal structures or nominal transparency, the cumulative effect of barriers is to lower effective transparency.

There seems to be tension between promoting accountability and legitimacy, through parliamentary and public participation in budgetary processes, and the achievement of fiscal discipline. An influential literature (e.g. von Hagen and Harden, 1995) emphasizes the importance of strong finance ministries to good outcomes in terms of the control of deficits and debt, exercising uncontested upwards transparency. Writing about Latin America, Alesina et al. (1999) and Stein et al. (1999) found benefits from strong hierarchical control of budgetary procedures and transparent fiscal information, in the form of lower deficits and debt. In a study of 19 OECD countries, Alt and Lassen (2006: 1403) confirmed that ‘a higher degree of fiscal transparency is associated with lower public debt and deficits’. In a panel study of US states, Alt and Lowry (2010: 379) found that ‘increased transparency dampens the negative effect of tax increases on the retention of incumbent governors’ but that ‘increased transparency leads to greater fiscal scale’. This study has the obvious strength of relating to sub-national governments operating within a relatively homogeneous institutional framework.

Writing specifically about Chile, Marcel and Tokman (2002) considered the additional mechanisms that are required to confer political legitimacy on a system of fiscal governance dominated by the finance ministry. It is paradoxical that much formal exposition of accountability processes starts from the assumption that the Executive wants to spend and the Legislature wants to economize, whereas the reverse is often the case. The political legitimacy of the Executive derives from the willingness of the Legislature to vote the resources necessary for government to operate. The huge symbolic importance of such legislative processes is often accompanied by actual procedures that have become enfeebled through neglect and are now largely formal. Party government has corroded authorization processes, so that the democratic credibility of the Legislature rests upon its capacity to exercise scrutiny. Yet the reality of oversight and scrutiny is often weak, with its backbone coming from the work of the Supreme Audit Institution. Fragile though it may be, acceptance of this downwards transparency differentiates those governments that depend on the democratic consent of the people from those that do not.

There are powerful theoretical arguments for simultaneous consideration of expenditure and revenue, so that both sides of the budget are seen together, as opposed to a situation in which expenditure and revenue are procedurally detached. In practice, this has proved difficult to achieve, in part because simultaneous consideration can overstretch administrative and scrutiny capacity. The paramount importance of seeing the link between expenditure and revenue has been stressed by research on developing countries (Bräutigam et al., 2008). Particularly in the context of countries rich in natural resources, rulers’ lack of dependence on citizens for taxation revenues has had traumatically damaging consequences for the quality and accountability of state activity and in terms of institutionalizing corruption. In this literature, taxation is argued to be the fundamental link between citizen and state, and the practical basis of the citizens’ accountability claims over those who presently control the state.

While individual countries have sought to resolve tensions about internal fiscal scrutiny in different ways, a marked development has been the growth of external surveillance of the fiscal affairs of individual countries. The causal factors behind this development are twofold.

First, globalization and greater economic interconnectedness have increased the spillovers from policy in one jurisdiction to others. A spirit of competitive collaboration, depending in part on institutionalized forms of peer review, has shaped the external scrutiny of member countries by the staffs of the OECD (economic surveys) and IMF (Article 4 consultations and Reports on Observance of Standards and Codes (ROSCs), with reference to the Code of Good Practices on Fiscal Transparency (IMF, 2007)). These sometimes have bite; some governments attempt to influence what is said about them and selectively leak ahead of publication, and published reports are selectively quoted by governments and opponents in domestic political debates.

Second, the development of the European Union has created another arena for external surveillance, but there are important differences that distinguish law-based surveillance (powers arise from European treaties) from membership based participation in IMF and OECD surveillance. The Stability and Growth Pact (SGP) applies to all 27 EU member countries but the context of application differs markedly in the case of the 17 members of the eurozone. This differentiation stems from the existence of a currency union in the absence of political union.

Sovereign governments taking instructions from supranational bodies raises questions about the legitimacy of international surveillance. How far these constrain domestic policy-making depends upon how these processes are perceived:

Are rules perceived to be applied even-handedly or can certain large countries operate outside the rules (e.g. France and Germany in relation to the SGP and certain countries in relation to ROSCs)?

These developments link to the depoliticization of public policy decisions (Flinders, 2005), whereby structures are established in which elected politicians defer either to pre-established rules or to independent experts buffered from everyday political life. In democratic countries, politicians can be ejected from office as a result not only of their mistakes but also because of unpopular well-founded decisions. This raises the question of the accountability mechanisms that apply to these independent experts. When experts, who advise or overrule politicians, make mistakes, who is to be held accountable? There might be independence from electoral politics and political decision-makers but not from particular modes of thought. Much judgement is embodied in economic forecasts and analyses which are to varying extents model-dependent. Experts may take credit for periods of (apparent) success in boom times, attributing these to superior conduct of policy, but then deflect blame when the economy goes into recession. Even when the proffered explanations are plausibly justified by evidence, the cycle of taking credit and deflecting blame brings discredit on expertise and on claims of benevolent independence.

4. Transparency about public expenditure

Having established both theoretical framework and context, this section analyses the transparency of public expenditure, with specific reference to the United Kingdom. Although constitutional and institutional arrangements vary across countries, there is a recognizable common core to public expenditure transparency.

The first step is to establish the meaning of transparency when applied to public expenditure, before turning attention to ‘intrinsic’ and ‘constructed’ barriers to such transparency. There are expositional advantages in presenting ‘intrinsic’ and ‘constructed’ as dichotomous, though these might alternatively be viewed along a spectrum that permits hybrid cases. Policy actors may deny that specific barriers are constructed to obstruct transparency, while some barriers may have constitutional or difficult-to-amend legislative origins. The distinction usefully structures the analysis and facilitates the development of remedies.

5. Transparency about surrogates for public expenditure

Spending money is not the only policy instrument available to governments: powers to legislate, regulate and tax are themselves fundamental. Surrogates for public expenditure require careful monitoring, particularly at times when there are pressures to restrain or cut what is counted as public expenditure. Moreover, the differential development of surrogates accentuates the difficulties of cross-country comparisons.

There are four main categories of surrogate:

off-budget expenditures, as when certain income sources and accumulated balances are not shown in budget documents (Kraan, 2004);

Surrogates are not necessarily inferior to public expenditure; the reason for identifying them is that – intentionally or accidentally – they may fall outside measurement systems, leaving their cost and distributional effects untracked. A case might be made for surrogates as preferable in particular circumstances for efficiency or equity reasons to ‘direct’ public expenditure, so the issue here is one of identification and scoring, not of instrument choice.

There are further issues of cultural and social acceptance. In Europe, there is no articulated opposition in principle to compulsory private car insurance for drivers; individuals have the choice of whether to drive, and there is concern about accident damage caused by uninsured drivers even in the presence of compulsion. Controversially, the Obama US health reforms involve compulsion that cannot be avoided and large coerced private expenditures that are not recorded as public expenditure.

This discussion of surrogates is a reminder of Wildavsky’s (1984) point that choices and processes in public expenditure budgeting are political, rather than technical, because they embody policy prioritization. Various motivations may lie behind the search for surrogates for public expenditure. Those who have a preference for near-market forms may believe that these mechanisms involve lower efficiency costs. Those who wish to see less income inequality might have recourse to surrogates in order to achieve concealed redistribution, notwithstanding widespread concerns that many actual tax expenditures are regressive in their distributional impact. Those who disagree with macroeconomic policy may see cut-price asset sales as less damaging to economic activity than cuts to current or capital expenditure. In order to achieve substantive policy goals, there may be a willingness to adopt non-transparent policy instruments.

6. Conclusion

The clearest conclusion is that fiscal transparency is fundamentally important but elusive. Claims made for transparency should be taken seriously but they should stand on their merits and should not be regarded as unanswerable. Attention has to be paid to context and to the uncomfortable relationship between transparency (the term likely to be used by those doing the watching) and surveillance (likely to be used by those being watched). Concerns about manipulative uses of transparency, even where justified, should not be used to under-value fiscal transparency.

Public accountability mechanisms require an appropriate fiscal transparency stance that determines the weight to be placed on transparency, relative to potentially competing values, and the desirable mix of fiscal transparency. On the latter, this article favours effective over nominal transparency and transparency in retrospect over transparency in real time.

The elusiveness of public expenditure transparency can be attributed in part to intricate technical material and limited public understanding of institutional structures. Fiscal transparency is situated within a complex pattern of institutionally structured incentives. Those who wish to improve fiscal transparency need to combine a realistic view of the behaviour of political decision-makers with an understanding of the pressures under which they operate. ‘Tough fiscal rules’, even in the form of constitutional amendments, may shortly be followed by the creation of arbitrage devices. Copying structures and mechanisms, without regard for context and habitat, is unlikely to be successful. However, some general principles can be articulated.

First, transparency is to be found in the disciplined release of information, without which claims to transparency will be undermined by manipulative practices such as plants, leaks and spin. A free-for-all damages effective transparency even if it appears to promote nominal transparency. On a technical level, the quality of some UK public expenditure documentation is exceptionally high, particularly the Treasury’s annual ‘Public Expenditure: Statistical Analyses’, and will help the United Kingdom towards high scores on international indices of budgetary transparency. This coexists with the intensive spinning of messages communicating misleading representations of UK public finances, thus damaging effective transparency. These practices confer such advantages on the Executive and on favoured media that there are grounds for scepticism on whether greater discipline in information release is achievable. The fact that this is tolerated is another reminder of the fundamental weakness in fiscal matters of the UK Parliament (Brazier and Ram, 2006).

Second, crucial ingredients of transparency are comprehensive coverage and a well-designed hierarchy of reporting documents, so that users can drill down from a synoptic overview to relevant entity detail where required. This has to be sensitive to constitutional arrangements and governmental practices; for example, in the United Kingdom this should be for the whole of the public sector, whereas in Australia documentation is more likely to relate to whole of jurisdiction (e.g. Commonwealth, or State of Victoria). The core idea is that the whole picture is mapped; this prevents unmapped recourse to off-budget expenditure (an endemic problem in Central and Eastern Europe in the years following the collapse of communism) and off-balance sheet assets (one of the attractions of PPPs to governments). The goal should be financial statements on an accruals basis, consolidation of entities, and full disclosure of any remaining off-budget and off-balance sheet activity. Accounting policies should be justifiable in terms of IFRS or the pronouncements of the International Public Sector Accounting Standards Board and there should be reconciliations between financial reporting and national accounts numbers. There is no contradiction between making the main documents as user-friendly as possible and ensuring that necessary technical material is accessible to those wishing to delve further. These demanding requirements have been achieved by best-practice countries and are achievable in the medium term by other industrialized democracies. For many other countries these will be aspirational targets to which a path has to be navigated.

Third, issues of capacity are closely linked to the structuring of political incentives. Effective transparency has to be a two-way relationship between those rendering the account and potential users of that account, here using ‘account’ in a generic and not exclusively financial sense. Constructive challenge is fundamental to the transparency process as, otherwise, the transparency provider is likely to lose interest or appreciate the opportunities for obfuscation. Compared with Executives, Legislatures outside the United States are often under-resourced, a situation compounded by many legislators seeing themselves as ‘ministers in waiting’. Various debate-suppressing tactics have been predominant at the Westminster Parliament, including the Treasury’s frequent practice of publishing Spending Reviews in July shortly before the long summer recess. Great importance attaches to the role of individuals with commitment and resilience (Murray, 2008); perhaps the greatest danger among legislators is to indulge in self-congratulation about scrutiny processes while the Executive knowingly looks on.

Fourth, the existence of scrutiny capacity outside the Legislature is to be valued. In the United Kingdom, the Institute for Fiscal Studies and the National Institute of Economic and Social Research have played constructive roles as commentators and analysts; both of these bodies enjoy sufficient status and reputation for independence to survive periodic bruising from offended politicians and ministries. There is international interest in the role of fiscal councils, at some remove from government and ministers and enjoying ‘guarantees’ of independence (Hagemann, 2010). While information brokers have valuable roles in clarifying choices and testing expert judgements, the actual delegation of fiscal functions to experts raises the accountability issues mentioned earlier. The degree of independence enjoyed by bodies such as the Swedish Fiscal Council and the UK Office for Budget Responsibility depends more on the cultural and political context than on issues of formal design. The transparency benefits are likely to derive from putting analyses in the public domain which external users could not perform for themselves. This inability may result from restrictions on external access to data or from the finance ministry considering particular analyses as too politically sensitive to publish, or even to undertake.

Fifth, the role of public audit is vital to effective transparency, both in terms of the validation of financial reporting (financial certification audit, analogous to that in the private sector) and in terms of its broader role in relation to regularity (expenditure is in accordance with authorization), propriety (absence of fraud and corruption) and performance (Value for Money work is now central to the role of many Supreme Audit Institutions). Public audit should belong to the Legislature (or judiciary where that is tradition), but certainly not to the Executive. The expanded role of public audit brings it into potential conflict with both Executive (it may criticize performance) and Legislature (it may cast doubts on policies favoured by legislators). This points to the delicate balancing required of the Supreme Audit Institutions, especially in relation to (what governments portray as) policy choices rather than implementation. A danger to be avoided is to accept roles beyond their technical capability.

Transparency creates a conflict between process and substance. Some actors may give absolute priority to substantive public policy outcomes (e.g. structure and coverage of public healthcare), whereas others may give absolute priority to the fiscal autonomy of sub-national governments. Others are likely to sit uncomfortably in between, with reactions influenced partly by principle and partly by opportunism. The central message of this article is that directions and varieties of transparency need to be studied carefully both at a generic level and in the specific case of public expenditure. Transparency cannot provide answers to profound ideological and practical questions concerning the scope of the state as measured by the size of public expenditure relative to the economy. But effective transparency about public expenditure can improve the evidence base on which informed views may be founded.