Maintaining market principles: Government auditors, PPP equity sales and hegemony

Abstract The UK's government auditors, the National Audit Office (NAO), play a central role in the accountability relations surrounding government expenditure. Commonly portrayed as being independent, they carry out performance audits assessing value for money. To date, the emerging market for PPP equity transactions has attracted little attention. This paper explores that emerging market through a Gramscian framework utilising the concepts of ‘common sense’ and ‘good sense’, focusing on a dialogical analysis of a NAO report on the topic.


Introduction
The UK's government auditors, the National Audit Office (NAO), play a central role in the accountability relations surrounding government expenditure. Commonly perceived as being independent, the main framing of their performance auditing (as opposed to statutory annual financial audit) work since the 1980s has been the value for money audit, focused on economy, efficiency and effectiveness (Broadbent & Laughlin, 2003;Funnell, 2011). In practice, this framing has led to a focus on saving money for the public purse. As the Auditor and Comptroller General has stated, the National Audit Office (NAO) sees their role in "undertaking value for money (VFM) audit [as] contributing to more cost-effective government" (Morse, 2010, p. 323).
Over the same period, the adoption of New Public Management (NPM) (Lovell, 1996;Power, 1997) techniques and policies has included the Private Finance Initiative (PFI) (Broadbent & Laughlin, 1999) where new public infrastructure assets are designed, built, financed and operated by a private company (known as an Special Purpose Company, SPC). The newly constructed asset (school, hospital, etc.) is then leased to the public authority over a specified time period (usually 25-30 years). While there is an established, if not exhaustive, literature on accounting related issues in PFI/PPPs 1 [see Andon (2012) for a comprehensive review], to date there has been little attention paid to the returns (profits) made through the sale of equity in SPCs by private sector partners. In recent years, some studies have emerged claiming excessive profitmaking through this route (HM Treasury, 2012;Whitfield, 2010). In response to these claims, the National Audit Office (NAO) published a report -Equity investment in privately finance projects (NAO, 2012a) -based on their investigation into the returns being made in the emergent PPP equity market. Significantly, this report did not adopt a value for money framing, but instead developed a model that sought to explain and justify the return levels reported from the equity sales.
Given this background, we seek to address the question, what do the contents of the NAO (2012a) report tell us about the nature and role of government auditors?
There are a limited number of theoretical or empirical studies addressing this question (one exception being a special issue of Critical Perspectives on Accounting in 2003). Beyond the functionalist, 'common sense' (Gramsci, 1971) view of government auditors as independent fulfilling a statutory role, Radcliffe (2008Radcliffe ( , 2011 argues that public auditors utilise 'open secrets' to take account of the views of potential audiences and acceptable policy discourses in their reporting. While there are elements of an open secret at work in the NAO report [i.e., that PPP remains an unpopular policy among the UK population and lacks both theoretical and empirical evidence to support its continued pursuance (Hellowell & Pollock, 2009)], we argue that the report represents a different approach that is best understood within a Gramscian framework of hegemony coupled with a dialogical understanding of language (Bakhtin, 1981). We draw on Gramsci's distinction between 'common sense' and 'good sense' ideas and integrate this with monological and dialogical understandings of utterances on 'returns' in the report (Bakhtin, 1981;Smyth, 2012). This leads us to conclude that in the case of this report, the NAO has become an active participant in maintaining the hegemony of competitive market principles as the basis of resource allocation for public infrastructure projects. This paper is structured as follows: Section 2 discusses what we know about the nature and role of government auditors by utilising Gramsci's work on hegemony, political and civil society. Further, the section introduces the 'common/good sense' distinction from Gramsci's writings and integrates this with the work of Bakhtin (1981) on centripetal and centrifugal forces in a dialogical understanding of language. Section 3 extends and applies this theoretical framing to PPP equity transactions by explicating a heteroglossic account. Section 4 explains the data collection and analysis techniques employed, before Section 5 presents the findings of our common/good sense analysis. Section 6 of the paper includes a discussion of the findings and concludes with a short discussion referring back to the original research question concerning the nature and role of government auditors.

Government auditors, hegemony and dialogics
There has been little work that attempts to theorise the social position and role of government auditors (Radcliffe, 2008(Radcliffe, , 2011. There is a common sense view that government auditors are independent of the world of politics and policy, are constrained by statute and custom to focus on financial and performance audit rather than policy objectives (Funnell, 2011). This view of government auditors' independence is deeply ingrained in the literature. For example, Normanton argues that "an independent state audit . . . is probably the citizen's best defence against misuse (using the word in its broadest sense) of money taken from his pocket" (Normanton, 1966, p. 410). Funnell quotes Pois, who concludes that as government auditors' independence is so widely accepted, there is little point in academics gathering arguments to support the proposition (Funnell, 2011, p. 716).
An alternative view has developed through the era of new public management (NPM) and is encapsulated in Michael Power's work on the Audit Explosion (Power, 1994). Power (1994) argues that audit symbolises independent validation, rationality, visibility and the promise of control, to which Lovell points out that "these might be comforting images, but they are argued to be mere chimeras" (Lovell, 1996, p. 267). Power later develops the idea that "audit represents a form of pragmatic 'muddling through"' (Power, 1997, pp. 143-144). English (2003) illustrates an alternative impact of NPM on the state audit function in Victoria, Australia, where the auditor-general's power to conduct audits was transferred to a statutory body, allowing for competition to develop with private audit firms through a purchaser/provider split. Despite the claims that this move was justified on economic grounds, English concludes that policy was politically motivated. 2 In the UK, government audit functions have been portrayed as being involved with a 'struggle for supremacy' between different audit agencies and these agencies and central government (Bowerman, Humphrey, & Owen, 2003), or having a legitimising role for government policies (Broadbent & Laughlin, 2003). Significantly, for this paper, Broadbent and Laughlin's (2003) work also focuses on the role of the NAO in relation to the early development and implementation of the PFI/PPP policy. Specifically, they argue that "that the NAO acts as such a legitimator in the context of their published advice on the calculation of value for money in PFI projects" (Broadbent & Laughlin, 2003, p. 30). The arguments in this paper follow a similar pattern to those of Broadbent and Laughlin (2003), where the NAO appear to adopt a policy analysis position that seeks to legitimate an existing government policy. In our case, there is an additional dimension that locates this legitimation role in broader hegemonic processes supporting the superiority of competitive markets as a resource allocator for public infrastructure investment purposes.
A more recent debate on the nature of government audit offices was conducted by Radcliffe (2008) and Funnell (2011, with the debate turning on whether government auditors' findings are independent of their policy and political context or moulded by the surrounding discourses and the need to gain a degree of acceptance by their audience. In response to Radcliffe's proposition that "government auditors express audit findings with care so as to work within prevailing discourses" (Radcliffe, 2008, p. 99), Funnell argues for the independence of government auditors and that "they are expected not to question the priorities and objectives of government, only the manner in which these may be achieved" (Funnell, 2011, p. 715). While we agree with much of Radcliffe's analysis and can see how it applies to the report analysed in this paper, we felt there is a need to move beyond the "careful measure of what is feasible" (Radcliffe, 2011, p. 724) and locate our analysis in broader socio-political contexts. Therefore, this paper utilises a Gramscian inspired framework, incorporating dialogics, to analyse the manner in which audit findings are produced.

Hegemony and contradictory consciousness
Gramsci's work on hegemony (1971) is an established framework within the critical and interdisciplinary accounting research tradition (Cooper, 1995;Lee & Cassell, 2008;Goddard, 2002;Spence, 2009). However, the key concepts from Gramsci's work used in this paper are his under-utilised (in the accounting literature) duality of 'common sense' and 'good sense'. This duality allows for an integration of Gramsci's conception of hegemony with Bakhtin's (1981) work on dialogics. Gramsci's idea of 'contradictory consciousness' (Gramsci, 1971, p. 333) is given life in Bakhtin's conception of a contradictory and stratified language (Bakhtin, 1981), as is argued below.
Gramsci was concerned with understanding how ruling elites maintain and reproduce their social position. He argued that they employ a dual strategy of coercion and consent. In the final analysis the state, with its monopoly of legitimate force in society (e.g. the police force) and the ability to the deny rights to its citizens (e.g. the courts and justice system), expresses the ruling elites strategy of coercion. The strategy of consent is more complicated relying on a combination of ideas with social and physical structures. Gramsci sees civil society as an essential element contributing to the ruling elites' hegemonic grip on society. In an echo of Marx and Engel's phrase -"The history of all hitherto existing society is the history of class struggles" (Marx & Engels, 1996, p. 3) -Gramsci outlines the relationship between the state and civil society: "The historical unity of the ruling classes is realised in the State, and their history is essentially the history of States and of groups of States. But it would be wrong to think that this unity is simply juridical and political . . . the fundamental historical unity, concretely, results from the organic relations between State or political society and 'civil society"' (Gramsci, 1971, p. 52).
This formulation is an outcome of the dialectical analysis employed by Gramsci, "the theory of the 'dual perspective"' (Gramsci, 1971, p. 170). Thus, building on "Guicciardini's assertion that two things are absolutely necessary for the life of a State: arms and religion" (Gramsci, 1971, p. 170), Gramsci formulates this duality in other forms such as force and consent, coercion and persuasion, political society and civil society, violence and fraud. These dualities are not rigid, but are context (socio-historical and economic) specific 3 and furthermore are dynamic with emergent powers, out of which a contestation develops between each and the context they operate within. 4 Applying this schema to government auditors, the NAO is positioned within the state/political society. It is independent from government departments, but relies upon statute to carry out its activities. However, it is not the case that government auditors are part of a coercive strategy to sustain the ruling elite. Further, their actions are portrayed as acting on behalf of the public -"helping the nation to spend wisely" (NAO, 2012a). Therefore, we argue that the NAO plays an important role in manufacturing the consent of civil society to the actions of the government. In this way, government auditors are not part of civil society, but act within it. Building on the dialectical conception of duality, Gramsci analyses the consciousness of the "active man-in-the-mass", where "his theoretical consciousness can indeed be historically in opposition to his activity. One might almost say that he has two theoretical consciousnesses (or one contradictory consciousness)" (Gramsci, 1971, p. 333). Importantly, Gramsci does not restrict this contradictory conscious to the individual, but also applies it to social groups and social strata; thus, "social groups which in some ways express the most developed modernity, lag behind in other respects, given their social position" (Gramsci, 1971, pp. 324-325).

Common sense, good sense and dialogics
The precise formulation that Gramsci uses to capture the two elements of contradictory consciousness is 'common sense' and 'good sense', where these elements are dialectically related in a whole, so that "[t]wo conflicting conceptions of the world usually exist, the one drawn from the official notions of the rulers, the other derived from an oppressed people's practical, although mediated, experience of social reality" (Cooper, 1995, p. 185). In this paper, common sense ideas are represented by the adoption and acceptance of market relations in equity transactions and profit-making on such transactions, as a positive development. In contrast, good sense ideas [or "new ways of seeing" as Cooper (1995, p. 186) formulates] are represented by challenges to the appropriateness of such transactions involving public assets such as schools and hospitals. Gramsci stresses that common sense/good sense applies to "every social stratum" (Gramsci, 1971, p. 326) and has its own historical sedimentation that influences the present. In true dialectical-style, Gramsci argues that good sense comes from within common sense, where there is a "healthy nucleus that exists in 'common sense' the part of it which can be called 'good sense"' (Gramsci, 1971, p. 328).
Dialogics have been intermittently used to analyse accounting information. The version of dialogics applied here is based on Bakhtin's (1981) original work in contrast to other approaches, such as Bebbington, Brown, Frame, and Thomson (2007) who draw on Freire's (1972) work to find a progressive engagement for social and environmental accounting. 5 Utilising a post-structuralist perspective, Macintosh relies heavily on Bakhtin's formulation of heteroglossia in his work (Macintosh, 2002). More recently, dialogics have been coupled with pluralism and agonistic democratic ideas (Brown, 2009) in an attempt to develop new and alternative accounting information systems (Blackburn, Brown, Dillard, & Hooper, 2014;Dillard & Yuthas, 2013). In contrast, the use of dialogics in this paper is on how currently existing accounting information interacts with existing ideas -common sense and good sense -in a hegemonic manner. In this respect Barker's (2006) dialogical analysis of a community consultation over the closure of a local hospital is more closely related to the approach in this paper. We argue that the same dialectical method used by Gramsci is present in the concepts of heteroglossia 6 and utterance. For example, Bakhtin (1981) advances a model that sees language as tension-filled and stratified, interacting with the concrete socio-historical environment (heteroglossia). Heteroglossia: "Represents the co-existence of socio-ideological contradictions between the present and the past, between differing epochs of the past, between differing socio-ideological groups in the present, between tendencies, schools, circles and so forth, all given a bodily form" (Bakhtin, 1981, p. 291).
In Bakhtin's work, there is an emphasis on contradiction, opposing social forces and change over differing time periods. However, Bakhtin's analysis goes further.
First, language is stratified by real, concrete social forces; for example, the manner in which the term "civil society" is used by advocates and opponents of neoliberalism and globalisation (Lavalette & Ferguson, 2007). Second, an episode of language -an utterance -enters into a dialogised heteroglossia, so that it does not relate to its object in a simple single manner, but interacts with a changing environment of alien words and themes about the same object. An example of this is the term an "unqualified audit report". The use of the word "unqualified" in this context is contrary to the common understanding of lacking in qualifications or measures of quality. In addition, once this (mis)perception is explained, a second level of issues arises over what an unqualified report signifies, with the audit profession restricting its meaning to an opinion based on audit procedures aimed at assessing whether the financial statements are materially correct, but not necessarily free of fraud. This restriction again contrasts with the common sense perceptions that it is the auditors' responsibility to find fraud, the financial statements are accurate and provide a truthful representation. The differences in these perceptions, known as the "expectations gap", reflect a strategy employed by the auditing profession to defend its social position following cases of audit failure (Sikka, Puxty, Willmott, & Cooper, 1998).
Third, the foregoing leads to both centrifugal (separating) and centripetal (unifying) forces at work in any utterance. Centripetal forces seek to unify a single meaning for the utterance, to make it monologic. In contrast, the centrifugal forces counter with multiple meanings, based on dialogised interactions. Thus, "Every utterance participates in the "unitary language" (in its centripetal forces and tendencies) and at the same time partakes of social and historical heteroglossia (the centrifugal, stratifying forces) . . . It is possible to give a concrete and detailed analysis of any utterance, once having exposed it as a contradiction-ridden, tension-filled unity of two embattled tendencies in the life of language" (Bakhtin, 1981, p. 272).
It is at this point that Gramsci's common sense/good sense duality can be applied to Bakhtin's work. In capitalist society, common sense ideas act as centripetal forces attempting to make an utterance monologic. Alternatively, good sense ideas act as centrifugal forces attempting to make an utterance dialogic. Applying this framework to the utterances in the NAO report on equity transactions in PPP projects, we see an attempt to justify the excessive profit making in the case study examples in a monological manner. However, there are centrifugal forces in the heteroglossia surrounding private sector profit-making in PPPs that counter. Further, there is a limitation to the monological explanation with the NAO admitting they are unable to explain all of the profits made on the equity transactions analysed.
These issues are explored further in the data and discussion below. First, it is necessary to establish the nature of the PPP projects, the emergent equity transactions market and the surrounding heteroglossia of this long term government policy.

PPP equity sales
There is a well-established field of accounting-related research into the PPPs [see Andon (2012)]. Rather than repeat a summary of this work, this section will briefly explain the nature of equity sales in PPP projects and the establishment of a secondary market. This will be followed by an exposition of the heteroglossia surrounding PPPs in Britain drawn from academic work and media reports, leading to a focus on the concern of this paper, the profit-making on equity sales in PPP projects.
PPP projects are delivered through a special purpose company (SPC), which contracts with the public sector organisation (e.g., an NHS Trust or local authority) to design, build, finance and maintain the hospital, school or other public asset. The SPC is highly geared, with up to 90 per cent of the total project finance in the form of senior debt from a bank or financial institution. The remaining finance is provided by the private sector partners who form the SPC (including the construction company and facilities management company). This balance is made up of equity (share capital) and sub-ordinated debt (loans from the private sector partners of the SPC). It is this equity 7 that is the focus of the analysis below, as it has given rise to the development of a secondary market (NAO, 2012a). As the National Audit Office outline, investors in PPP equity: ". . . have typically been either contractors, who also provide services under the contract, or financial institutions. Some investors are interested in a long-term involvement with a project. However, many of those investing in the project at its start, known as primary investors, will sell their shares soon after the new asset has been delivered in order to fund new projects. These primary investors sell their shares to secondary investors who want a long-term stable income from mature projects" (NAO, 2012a, p. 4).
There is little literature addressing the emergence of this secondary market. The sale by Serco of their holding in the Defence Helicopter Flying School PFI in June 1998 is the earliest case of an UK PPP/PFI equity transaction (ESSU, 2012).
Significantly, Serco reported a £4.6 million profit on this disposal in their financial statements for 1998. Whitfield (2010) first outlined the scale of PPP equity sales transactions and related profit-making; this was followed by an ESSU research report in January 2011, resulting in evidence given to House of Commons Public Accounts Committee and Treasury Committee investigations in 2011 and a BBC Radio 4 programme in June 2011.
This work led the NAO to investigate the practice of equity sales in PPP projects (NAO, 2012a). The NAO report concluded that the development of a secondary market in PPP equity sales is a positive step that could lead to a reduction in the return (cost) of equity in SPCs, as it gives some confidence to the primary investors that they have an exit opportunity (House of Lords, 2010). However, the HM Treasury review leading to the revised PPP proposals admitted there have been windfall gains on equity transactions with third parties (HM Treasury, 2012). We will return to the level of profit making by PPP private investors below.

Heteroglossia: PPP/PFIs and investor returns
Andon (2012) takes and adapts the research agenda developed by Laughlin (1999, 2004) to review the accounting related research that has been completed up to 2012. This research agenda covers the underlying nature and rationale for PPP; processes aiding decisions to undertake and ex-post evaluations of, PPP projects; the merit and worth, regulation and guidance of PPP projects (Andon, 2012, pp. 877-878). Andon concludes his review by stating: "The extant literature has critiqued motivations, discourses and influences that have shaped PPP developments. It has made important contributions to explicating the deficiencies of and values associated with financial calculations used to assess VFM and manage risks for proposed PPP schemes. Extant work has highlighted the scant development of ex post evaluations of PPP schemes, and has offered advice on how this could be remedied. The literature has drawn attention to the conflicting evidence, difficulties and ambiguities in substantiating the merit and worth of PPP schemes. It has also flagged important challenges in making PPPs appropriately transparent and accountable" (Andon, 2012, pp. 900-902). This is a comprehensive list of the failings, problems and limitations of PPPs found in the literature. More specifically, Hellowell and Pollock argue that: "Investment through PFI cannot, in and of itself, lead to fiscal savings; and the economic case for the method is based on false assumptions and misleading evidence" (Hellowell & Pollock, 2009, p. s23). Significantly, in this context they locate the motivations for the enduring use of PPPs as lying in successive governments' desire to expand private sector activities in the economy and "is driven by a (related) ideological belief in the inherent superiority of private over public models of investment and service delivery" (Hellowell & Pollock, 2009, p. s23). This stance was most forcefully expressed by one New Labour 8 Minister: "When there is a limited amount of public-sector capital available, as there is, it's PFI or bust" (BBC News, 2001). We argue in this paper that this ideological belief is also expressed in the adoption of competitive market-based principles by successive governments as the only acceptable resource allocator for public funding.
In contrast, successive UK governments have argued that PPPs offer value for money for the taxpayer; that they offer a basis on which to transfer risks to the private sector contractors and financiers and enhance service delivery through access to the latest private sector techniques. This is despite the academic evidence challenging and/or contradicting each claim. Further, the governments' support for PPPs has been challenged beyond the walls of the academy, through stories in the mass media and reports published by civil society organisations, such as the trade union Unison. For example, in June 2015, The Independent newspaper in the UK ran a story with the headline: "Britain's first PFI privately funded NHS hospital is a 'major' fire safety risk, say fire fighters" (Stone, 2015), which reported deficiencies in the design and construction of the hospital. A report commissioned by Unison into whether there had actually been an evaluation of the relationship between risk transfer and higher costs of private sector finance (in comparison to government debt) found a lack of evidence to show the work had been completed (Pollock & Price, 2004). Unison also mobilised the work of the NAO to support their opposition to PPPs. In 2005, commenting on the publication of an NAO report into the debt refinancing at Darent Valley hospital, Unison stated: 'The National Audit Office (NAO, 2005) report into PFI-funded Darent Valley Hospital, is proof positive that private companies are profiteering from PFI deals' (Unison, 2005). The effect of these challenges to the government policy has been evident in wider society, with a range of opinion polls showing either concern over the policy or outright opposition to it. For example, in 2007, voters in Scotland rated the following as the most important in a series of policy related statements: "Ensure that all state schools and hospitals are built and run by public bodies rather than private companies" (BBC News, 2007). A major area of concern for the public, as well as academics, is the large profits being made by private sector partners from PPPs (Demirag, Khadaroo, & Stapleton, 2015). These profits (often labelled using the more neutral term "returns") have come from a variety of sources, as explained below.
The focus of research on returns for PPP projects has thus far been based on ex ante, projected numbers. This is in the main a function of the relatively early stage of the policy implementation and project development. It was not until after the New Labour government came to power in 1997 that the PPP policy started to result in significant numbers of projects. For many researchers, the only (semi-)available financial information was drawn from the limited number of business plans of SPCs that became publicly available. Vecchi, Hellowell, and Gatti (2013), while analysing the projected Weighted Average Cost of Capital (WACC) and Internal Rate of Return (IRR) of 77 hospital PFIs in the UK, conclude that "the results confirm that the sponsors of the PFIs analysed have extracted a higher return than their cost of capital" (p. 255). There have, however, been a limited number of studies that have utilised ex post accounting information. Edwards, Shaoul, Stafford, and Arblaster (2004), Shaoul, Stafford, and Stapleton (2008) and Bain (2008) all utilise accounting information and an accounting-based analysis to calculate key ratios. Significantly, all three studies show that private finance is more expensive than the borrowing costs the government could and did achieve, even if Bain (2008) argues that the private finance premium is lower than that found by Edwards et al. (2004).
This section has a twofold purpose, first to explain the nature of PPP equity transactions and the development of a secondary market in their trading. Second, it has outlined the heteroglossic context covering PPP/PFI into which the NAO report analysed below entered. This is a context that has multiple and competing meanings when PPP/PFI is addressed, central to which is the question of whether excessive returns (profiteering) are being made. This question provides the basis for the analysis in Section 5. Beforehand, the next section sets out how the NAO (2012a) report was analysed with the focus on "return" utterances.

Method -data collection and analysis
The NAO report Equity investment in privately finance projects runs to 36 pages and includes a summary, three parts (covering The role of equity investors; The investors' management of risk and contribution to service delivery; and Returns to investors) and an appendix on overall methodology. There are three more appendices available at a different location on the Internet including, significantly, an appendix (No. 4) that "sets out the methodology behind our illustrative analysis of returns to primary investors" (NAO, 2012b, p. 2).
The dialogical analysis in this paper was developed in two ways, drawing on similar techniques to those used by Morales, Gendron, and Guénin-Paracini (2014). First, the NAO (2012a) report was read closely on multiple occasions. The initial close readings were used to familiarise ourselves with the contents of the report, paying particular attention to the aims of the study and how they were addressed. This led to us settling on the focus of this paper, the attempt to justify the excessive/unexplained profit-making occurring in the secondary PPP/PFI market. This is the framing the NAO adopt, rather than demonstrating whether value for money (from the public's perspective) has been established, something that the report specifically excludes from consideration: "As the scope of this report is specifically limited to the role of PFI equity, it does not deal with all the issues that would be relevant to assessing the value for money of each project. For example, Portsmouth Hospitals NHS Trust encountered budgetary difficulties, which in March 2011 contributed to its decision to close 100 of the Queen Alexandra Hospital's 1,200 beds" (NAO, 2012b, p. 5) This statement struck us as an important admission of the NAO's motivations for the report and hence, worthy of deeper, critical investigation. The subsequent close readings took place alongside the key moment analysis. In the main, they were used to confirm the findings that were emerging from that analysis (Morales et al., 2014). Second, an analysis based on a dialogical (Bakhtin, 1981;Sullivan, 2012) understanding of the nature of language was applied. Dialogics portrays language as tension-filled and stratified, interacting with the concrete socio-historical environment (heteroglossia).
"At any given moment . . . language is stratified not only into linguistic dialects in the strict sense of the word . . . but also . . . into languages that are socio-ideological: languages of social groups, "professional" and "generic" languages, languages of generations and so forth" (Bakhtin, 1981, pp. 271-272).
Sullivan's (2012) data analysis approach includes: "'Key moments' [as] an 'utterance' of significance. An utterance is a significant unit of meaning . . . and is defined by its readiness for a reply/reaction. As a unit of meaning, it can be of variable length" (Sullivan, 2012, p. 72).
Applying these ideas to the data collected, the central analysis technique used was to identify the key moments in the NAO report (2012). These key moments recognise the multiple voices that are present in the text (Brown, 2009), despite the attempts to present a monological justification of the profit-making in the secondary market. The NAO report was searched for key words (utterances) ("gain", "profit" and "return") 9 with the resultant key moments noted and highlighted. The results of this search generated nine utterances of "gain", four utterances of "profit" and 116 of "return". Given these responses, further detailed analysis was carried out on the return utterances identified, while the gain and profits utterances were consulted for confirmatory purposes. Table 1 gives some indication of how often "return" utterances were made in the report; but this gives little insight into the nature of the utterances in relation to common sense/good sense ideas. For example, 30 utterances relate to headings in the report or outlining the research method employed. Yet, contentious issues such as debt re-financing or capital recycling are rarely mentioned in relation to returns. Therefore, the reporting of the findings in the next section seeks to analyse the utterances in relation to their common sense/good sense basis, as illustrated in the analysis in Table 1. Each utterance was allocated as representing either common sense or good sense or both, on the basis of whether it complies with conceptions of competitive market principles (see Section 5). The analysis excludes those utterances concerning the Headings in the report. This is for two reasons; first, some headings were so limited in their communicative content (such as "Returns to investors" on pages 3 and 23) that we considered it inappropriate to classify them as either common sense or good sense. Second, we also knew that the substance of any potential communication in any heading would be picked up in our detailed analysis of the contents of each section. Therefore, we analysed one hundred "return utterances" with the results reported in the next section.
The next section sets out the findings of this analysis, culminating in a more detailed exploration of one utterance concerning the Queen Alexandra Hospital, Portsmouth project. This is one of three projects selected by the NAO in an attempt to explain all elements of the return made by a private sector partner when they sold their equity. Using competitive market principles as expressed through the risk-return model, they are unable to do this.

Findings
This section presents the results of our dialogical analysis (Smyth, 2012) utilising Gramsci's common sense/good sense distinction. The NAO report is a professional document in appearance. Its cover carries the official logo of the National Audit Office and key identifying information such as the title, date, reference number and which parliamentary session it was published in. The document is freely and publicly available on the Internet. Inside the front cover, there is the vision statement of the NAO: "Our vision is to help the nation spend wisely. We apply the unique perspective of public audit to help Parliament and government drive lasting improvement in public services" (NAO, 2012a). On page two of the report the NAO states: "This report gathers together the available information on PFI equity and combines it with further information we have collected from projects and investors. We use the information to set out a more extensive analysis of the risks and rewards that PFI investors have experienced" (NAO, 2012a, p. 2).
The remainder of this section sets out the common sense/good sense analysis of the categories of utterances in Table 1. First, we establish the basis on which this analysis is developed, the competitive market principles.

Competitive market principles
We argue that in the assumptions and aims adopted in Equity investments in privately finance projects, the NAO have moved beyond the open secrets strategy identified by Radcliffe (2008Radcliffe ( , 2011 or legitimising role (Broadbent & Laughlin, 2003), and is instead actively taking part in maintaining the ideological preference of the ruling elite for competitive markets as the basis for allocating resources in public infrastructure projects. Early in the report, the NAO state that one of their aims is to examine whether 'the returns, for investors, are transparent and reasonable, derived from contracts priced in line with market principles' (NAO, 2012a, p. 5). These market principles are not set out at any point, but from the analysis of the utterances on returns we take them to include: 1. The application of a risk-reward model drawn from finance theories. 2. A competitive procurement process. 3. A degree of transparency (within commercial confidentiality norms). 4. A strong legal and contractual basis. 5. An active public sector partner who monitors compliance with the contractual requirements.
In the analysis that follows, utterances were allocated to either common sense (those that support the above principles) or good sense, where the above principles are challenged or shown to be limited.

Findings -return levels
As is expected with a dialogical analysis, many of the utterances in this category contained both common sense and good sense forces. For example, the reporting of rates of return could be seen as supportive of the developing secondary market in equity transactions, giving a return to equity providers who take the ultimate risk of losing their capital (NAO, 2012a, p. 5) or could be considered evidence of profiteering by private sector partners at the expense of the public. For example: "Our analysis has shown that investors selling shares early have typically earned annualised returns between 15 and 30 per cent. In exceptional cases, returns have been higher (up to 60 per cent) or lower (as low as 5 per cent)" NAO (2012a, p. 7).
There were eight utterances on return levels that we categorised as common sense only. These utterances covered a variety of return related issues, but all supported competitive market principles. For example: "Whilst the majority of project deliveries have been managed effectively by investors, there have been some examples where investors lost money or made a lower return than was forecast in their bid" (NAO, 2012a, p. 18). We also found two utterances that represented good sense ideas where the NAO stated they could not explain all of the profits being made through equity sales by using the competitive market principles. For example: "The parts of the investors' returns which could not be fully explained were a relatively small amount -around £1.15 million per annum in total across the three projects -but they were equivalent to around 1.5 to 2.2 per cent of the authorities' payments and could be significant over the long term life of PFI projects" (NAO, 2012a, p. 7).

Risks and returns
The prevalence of the risk/return model, where "equity investors' returns are expected to be high relative to the senior debt lenders because they take on greater risk" (NAO, 2012a, p. 15), is an extension of the adoption of market principles as the NAO's guiding basis for analysing the secondary market. As this trade-off between risk and return is central to the operation of efficient markets, all but one of the utterances in this category supported common sense ideas. The one exception was where the NAO recognised that when all aspects of a competitive market are in operation, ". . . investors do not have to demonstrate that their returns are reasonable for the specific risks of the project" (NAO, 2012a, p. 23). Here the government auditors admit that the link between risk and return may have broken down, and in the process, good sense ideas (for example, excessive profit-making) may be at work. This point is further emphasized in another utterance that we categorised as containing both common and good sense ideas: "However, there is a reputation risk to a private finance programme when investors are perceived to be earning high returns from government projects" (NAO, 2012a, p. 5). This utterance supports the finding in Demirag et al. (2015) where some equity financiers have ruled out an early exit from PPP projects through equity sales because of the potential damage to their reputation. However, the utterance also recognises the heteroglossia around perceived profiteering by private sector partners. Consistent with the risk-return framing adopted by the NAO, there is an attempt to translate potential damage to reputation into the language (and calculations) of risk.

Research method employed
As would be expected given the assumptions made by the NAO in preparing their report, a significant majority of the utterances (9 out of 14) concerning the research method carried common sense ideas only. Three out of the four utterances that contained both common and good sense ideas all alluded to the limitations of competitive market principles, in that those principles may not be operating in practice as the theory states they should. However, the most significant utterances under this heading concern how the NAO went about deciding what projects were to be included in their analysis. As one utterance illustrates: "An investor informed us about a purchase in December 2005 in which the rate of return was 12.1 per cent, which was considerably greater than rates of return linked to other sales in the same month. We did not include this rate of return . . . because it was an outlier and we had no knowledge of the project(s) included in the sale" (NAO, 2012a, p. 31).
This utterance clearly shows the tension-filled, competing common and good sense ideas in the report. On one hand, it would appear to be common sense to exclude an "outlier" that was difficult to verify. However, from a good sense perspective, this could be seen as an attempt to manipulate the numbers, once again with the aim of trying to justify the returns being made in the secondary market. This perspective is further supported by the one good sense utterance where the NAO, having failed to justify the total return (profit) made on the equity sales in their three case studies, attempt to downplay the significance by saying the "results are not conclusions on the value for money of the three projects and should not be taken to indicate experience across the whole population of PFI project sales" (NAO, 2012a, p. 34).

Market operations
This category contains utterances about the actual operation of the PPP market to date. We identified utterances concerning inefficiencies, monitoring and competition. As with the other categories a compelling majority of utterances contained common sense ideas; thus, "competition has generally created an expected return to equity of between 12 to 15 per cent at the point contracts are signed" (NAO, 2012a, p. 6). However, the report also contains utterances about market inefficiencies such as poor monitoring by public sector partners and a lack of capacity to challenge projected returns: "Authorities have, generally, not been equipped to challenge investors' proposed returns rigorously and may require better support to do so" (NAO, 2012a, p. 7). And: "In general, the Treasury and Authorities have relied on competition to seek efficient pricing of the contract, without systematic information to prove the pricing of equity is optimal" (NAO, 2012a, p. 23). Significantly, given the dialogical analysis below of the Queen Alexandra Hospital PPP, the report also states, "In 2002, PricewaterhouseCoopers published a report, commissioned by the Office of Government Commerce, which concluded that there had been inefficient pricing built into past equity rates of return" (NAO, 2012a, p. 15).
This utterance contains good sense ideas which challenge the centripetal forces that market principles work in practice and also feeds into the heteroglossia that private sector partners are making excessive profits from PPP projects out of public funds.

Justifications/explanations
It is no surprise, given the competitive market principles basis adopted by the NAO, that the utterances in this category all contain common sense ideas. These are ideas whose justification rely on the risk-return model. There is one utterance that contains both common sense and good sense ideas where the NAO recognise that "[s]ome examples of high investor returns have attracted adverse publicity" (NAO, 2012a, p. 4). The common sense idea of an attractive return available for potential market participants, clashes with the good sense excessive profit-making narrative. Significantly, the NAO include a quote/response from HM Treasury (HMT) on their findings. In this utterance, HMT seek to justify the unexplained excessive profits identified in the report by broadening the focus of the analysis such that readers need ". . . to take into account a wider range of issues that together contribute to the overall economics of a transaction, rather than merely looking at equity returns on their own" (NAO, 2012a, p. 8). It should be stated that despite this, the NAO "could not explain a proportion of the returns earned by investors" (NAO, 2012a, p. 7). Further, HMT do not state what other economic issues should be included beyond the risk-return model employed by the NAO. a Includes accrued profit at the secondary investor's rate of return -i.e., those relating to operational risks (NAO, 2012a, p. 34). b These risks include only those risks that primary investors are exposed to, but secondary investors are not (NAO, 2012a, p. 34).

Lack of transparency
This category included the most number of utterances containing good sense ideas (i.e., those that challenge the appropriateness of competitive market principles as the basis on which to allocate capital infrastructure resources). In the report's Summary, the NAO state: "Publicly available information on investors' risks and rewards from private finance projects is limited" (NAO, 2012a, p. 4). This admission continues an established (good sense) critique of PPP projects (Shaoul, 2006). With regard to PPP equity transactions, the report does shed some light on why this is the case. First, there is the recognition that there is no legal requirement on behalf of the vendors to disclose any details of the sale (NAO, 2012a, p. 15). Second, as the sale of equity is between two private entities (for example, the SPC and an infrastructure investment fund) the NAO claims it is beyond their remit to investigate. However, "in the absence of systematic information more detailed analysis of project returns can help to assess whether equity pricing is reasonable" (NAO, 2012a, p. 7). This utterance contains good sense ideas (the lack of available information) and illustrates the common sense ideas that through detailed analysis (making assumptions and building models), it is still possible to assess reasonableness (based on the risk-return trade-off).

Other -recommendations, debt re-financing, capital recycling
All the utterances in these categories exhibited common sense ideas. For example, the recommendations are all aimed at enhancing and developing the competitive market forces. However, the three debt re-financing utterances also contained good sense ideas, as they all reflect previous research showing examples of excessive profits made by private sector partners through this activity. The capital recycling utterances all made similar common sense points: "Share sales have enabled primary investors to release their capital and fund new projects, thereby also accelerating their returns" (NAO, 2012a, p. 7). It must be stated that at this stage, there is no evidence that such recycling has resulted in funds being made available for new projects that would not have been possible if the secondary market did not exist. This raises the concern that the motivations for entering the secondary market are based on accelerating returns (profit-taking).

Heteroglossic/dialogical example
By way of drawing this analysis to a conclusion, we focus on one particular utterance that the NAO highlight. In addition to an overall analysis of 99 equity transactions, the NAO look in detail at three equity transactions, utilising the competitive market principles outlined above, seeking to explain the totality of the returns (profits) reported. Table 2  Based on these calculations, the NAO estimate that 2 per cent of the annual unitary charge is unexplained. The manner in which the calculations are constructed is consistent with the aims of the report [i.e., to explain the returns (profits) made on a PPP equity transaction]. The extract shows the attempt to establish a monological understanding of the returns being made in the secondary market. One can easily imagine that if the NAO had been able to establish an explanation for all of the return (using a risk-return model), this would justify the emergence of the secondary market and so would be the end of the matter. Instead, they state: "These illustrations do not represent a conclusion on the value for money of those projects and should not be taken as indicative of similar questions in other projects" (NAO, 2012a, p. 7).
Of course, even if the NAO had been able to explain all of the return (profits), dialogical explanations would still be possible, such as those that critique and expose the limitations of competitive market principles. However, the models and assumptions the NAO have adopted are unable to explain the totality of the returns made, thus exposing the limitations of monological (common sense) understanding and creating the basis for dialogical (good sense) understandings. One dialogical explanation is to apply basic financial accounting rules and calculate a profit/loss on disposal. In the case of Queen Alexandra Hospital, this results in a gain (or profit) on disposal of £24.1 million (31.3-7.2). In turn, this new utterance allows for a further insight. It is our contention that as the £24.1 million gain is based on actual transactions, it holds more credibility than trying to explain away component parts of the gain using estimated risks and other valuations. After all, Carillion was able to use that gain to pay dividends (the dividend paid in 2010 was £61.4 million 10 ), representing a leakage of public funds (Shaoul et al., 2008) to private shareholders.
This section has set out the findings of our analysis of return utterances in the NAO report, identifying the common/good sense ideas within the utterances and culminating in an analysis of one prominent utterance, the Portsmouth hospital project. In this analysis, we have been able to illustrate the tension-filled nature of the language used by the NAO. The next section discusses these findings and draws a conclusion on what they tell us about the nature and role of government auditors.

Discussion and conclusion
The analysis in the previous section cannot be relied upon for robust and rigorous findings on the basis of a simple count of utterances; this would contravene the theoretical framing using dialogics and hegemony. It is noteworthy that in each utterance category, common sense ideas were more prevalent than good sense ones. Instead, we need to locate the utterances in the broader heteroglossia surrounding PPP projects and the role of government auditors. At the start of the previous section, we introduced a list of five competitive market principles including the risk-return model. This was the second most prevalent category, with common sense ideas to the fore. However, while these ideas were dominant, they were not totalising as there were good sense ideas in some utterances, especially when the actual return levels were presented in the report. This is especially the case when the unexplained (excessive) returns were reported, thereby, highlighting the limits to monological utterances in that they fail to explain the totality of social phenomena (in this case the profit-making in the PPP secondary market). Using the common sense ideas of competitive market principles, for the NAO the residual amount in the sales price is 'unexplained' (e.g., the £3.4 million in the Portsmouth hospital example). Because they have refused to move beyond the dominant ideology of competitive market principles, the NAO have reached the limit of their ability to analyse and explain what is actually occurring in the PPP equity market.
However, by adopting a dialogical approach to the 'unexplained' residual amount, alternative understandings can be developed. One possibility is that the examples analysed by the NAO are not representative of the whole market. The NAO hint at this when they state that the examples "should not be taken as indicative of similar questions in other projects" (NAO, 2012a, p. 7). This is an attempt to salvage the common sense competitive market principles in the abstract, in that the principles are fundamentally appropriate but are just not working in these particular cases. There are other utterances that support this interpretation when the NAO refer to the lack of transparency in the market and the poor monitoring record of public sector partners, both of which are considered necessary for an effective, competitive market to operate. An alternative and more challenging dialogical explanation is that profiteering is occurring by private sector partners. It must be remembered at this point that for most PPP projects, the only source of income comes direct from public funding. The NAO recognise this when they state that as part of the motivations for their report, if there is an unexplained residual amount, it would indicate the public sector may be paying more than it should be (NAO, 2012a, p. 23). Further, the heteroglossia surrounding private sector profits from PPP projects enhances the argument that profiteering is occurring. For example, Edwards et al. (2004), Shaoul et al. (2008), Bain (2008) and Whitfield (2012) all indicate that there is an additional premium to using PPP (as opposed to traditional government funded procurement) even if there is some disagreement about how much the premium may be. It is our proposition that in line with Radcliffe's (2008Radcliffe's ( , 2011 open secrets, the NAO are aware of the excessive profit-making, but know that they cannot openly report it. Instead, we are left with 'unexplained residual amounts'. Our analysis does not terminate at this point, as by utilising a Gramscian framework we are able to contribute more to the debate about the nature and role of government auditors. Existing literature portrays them as either independent professionals pursuing ('muddling through') a key activity in democratic states; as the purveyors of open secrets, where consideration of current political narratives and reactions from potential audiences is an important influence; or as a legitimating process for government control. In contrast, our conclusion is that in this report, the NAO has played an active hegemonic role, advocating the ideology of competitive markets as the only acceptable resource allocator for public infrastructure projects. This conclusion is further enhanced by the deliberate choice of the NAO not to consider value for money as the basis on which to analyse the transactions as highlighted in the Queen Alexandra Hospital example. This conclusion is based on locating government auditors as part of the state (political society), but with a consent-manufacturing role in extending hegemony in civil society. This hegemonic role is usually limited to evaluating government activities using a value for money framing and on behalf of the public. In this report, by abandoning that framing and adopting an analytical framework based on competitive market principles, the NAO have played a conscious role in maintaining the hegemony of successive governments.
The major limitation to this conclusion concerns the focus on only one report, which, while we have pursued an appropriate and robust research design and theoretical framing for the analysis, means it is not possible to generalise more broadly about the nature and role of government auditors; this is the task of the research to come. Other limitations include that the NAO failed to examine the wider and longer-term implications of PPP equity sales, assess the impact of the increased concentration of PPP assets in infrastructure funds and joint ventures. Nor did it investigate the increasing use of tax havens and quantify the potential loss of government tax revenue. A subsequent Public Accounts Committee hearing did not challenge the NAO's methodology or analysis (Public Accounts Committee, 2012). All these concerns lead to a potentially fruitful future research agenda that will enhance that already developed by Broadbent and Laughlin (1999).
Finally, on a more positive note, the linking of the dynamic, tension-filled version of dialogics in this paper with Gramsci's common/good sense ideas opens the space for the work completed by academics that are critical of government policy on PPPs to have a real impact through establishing and enhancing centrifugal forces and a dialogised heteroglossia. In the process, the good sense ideas can be picked up and amplified by social movements and civil society organisations (such as trade unions). This seems to us to be a positive potential development in challenging monological and hegemonic ideas prevalent among the UK and other governments on PPP projects.