The paradox of accounting for cultural heritage: a longitudinal study on the financial reporting of heritage assets of major Australian public cultural institutions (1992–2019)

Purpose – The monetary valuation of cultural heritage of a selection of 16 major public, not-for-profit Australian cultural institutions is examined over a period of almost three decades (1992–2019) to understand how they have responded to the paradoxical tensions of heritage valuation for financial reporting purposes. Design/methodology/approach – Accounting for cultural heritage is an intrinsically paradoxical practice; it involves a conflict of two opposite ways of attributing value: the traditional accounting and the heritage professionals (or curatorial) approaches. In analysing the annual reports and other documentary sources through qualitative content analysis, the study explores how different actors responded to the conceptual and technical contradictions posed by the monetary valuation of “heritage assets”, the accounting phraseology of accounting standards. Findings – Four phases emerge from the analysis undertaken of the empirical material, each characterised by a distinctive nature of the paradox, the institutional responses discerned and the outcomes. Although a persisting heterogeneity in the practice of accounting for cultural heritage is evident, responses by cultural institutions are shown to have minimised, so far, the negative impacts of monetary valuation in terms of commercialisation of deaccessioning decisions and distorted accountability. Originality/value – In applying the theoretical lens of paradox theory in the context of the financial reporting of heritage, as assets, the study enhances an understanding of the challenges and responses by major public cultural institutions in a country that has led this development globally, providing insights to accounting standard setters arising from the accounting practices observed.


Introduction
More than 80 years ago, in 1937, T. R. Adam wrote in regards to the Civic Value of Museums, and provided counsel about the monetary values of museums: To attempt an estimate of the money value of the contents of our museums would be an intellectual vulgarism. . . . collections created to illustrate the achievement of man's (sic) hand and eye lie outside the field of exchange . . . The concepts of a museum as something that can be bought with money is common but misleading (Adam, 1937, p. 2).
For decades and centuries, cultural institutions (such as museums and art galleries) have sought to achieve a purpose similar to that as described by Adam (1937) above and this purpose is represented as such in their mission statements (see, for example, Carnegie and Wolnizer, 1996). Despite existing for the benefit of the public and without profit-making aims or intentions, the monetary valuation of collections for financial reporting purposes is firmly on the agenda for the international community of public cultural institutions (De Wolf et al., 2020).
Set in Australia, this study on the monetary valuation of the cultural, heritage and scientific collections of not-for-profit, public cultural institutions for financial reporting is a contribution to a long and ongoing discussion stemming from world views held within two professional communitiesprofessional accounting and heritage experts [1]. This discussion is focussed on the conflict between accounting and heritage professional (or curatorial) approaches to the valuation of the collections of these social institutions. Specifically, professional accountants aspire to attribute financial values to collections, whereas heritage experts are concerned with the non-financial values of the collections to society. The accounting profession views such collections as assets, specifically known as "heritage assets", for financial reporting purposes such as in the Consultation Paper (CP) issued in 2017 by the International Public Sector Accounting Standards Board (IPSASB) on the "Financial Reporting for Heritage in the Public Sector". The heritage professionals, on the other hand, expressly acknowledge the broad-based cultural, heritage and scientific values of prized public collections as well as their educative values, which each represent non-financial values in the unique contexts in which such museums operate Wolnizer, 1995, 1996).
This longitudinal study in the interpretivist research tradition is intended to inform the interdisciplinary discussion and debate on this controversial topic (Carnegie and Wolnizer, 1995, 1997a, 1997bChristensen and Mohr, 1999;Barton, 2000Barton, , 2005West, 2003, 2005;Hooper et al., 2005;Ellwood and Greenwood, 2016;Woon et al., 2019;Anessi-Pessina et al., 2019). Rather than to necessarily add to the theoretical or conceptual arguments of one-side or the other, it is concerned with providing over-arching evidence of financial reporting practices across a continuous period of 28 years from 1992 to 2019. By means of exploring the actual accounting practices of a sample of Australian not-for-profit, public cultural institutions across this time span, the objective of this study is to investigate the evolving pattern of responses to the paradox of applying accounting logic to the valuation of culturally relevant artefacts for financial reporting purposes. The study's key research question is stated as: "What practices and trends in the financial reporting of collections of selected Australian cultural institutions have occurred given the diversity of views, and related controversy, between accounting and curational approaches to valuing the collections? ". The paper employs a processual approach to analyse responses to paradoxical tensions (Smith and Lewis, 2011;Jay, 2013;Jarzabkowski et al., 2013;Hargrave and Van de Ven, 2017).
The findings are expected to be of interest to both sides of this long-running and extensive argument by providing a unique perspective of the results relating to this accounting practice in four phases across the study period of 28 years. From an investigation of accounting policy and practices based on disclosures in annual reports, insights may be derived for reviewing and adopting future policy. Such policy may ensure adequate care and protection is afforded to these prized and unique public resources in the contexts in which they are deployed for preservation and conservation into perpetuity. It is contended that accounting for cultural heritage needs to be consistent with the mission statements of cultural institutions and should not adversely affect programmes for the conservation of collections. Further, accounting in any context should satisfy the cost-benefit criterion.
The study endeavours to make three key contributions. First, evidence relating to the practice of accounting for heritage assets in financial reports comes from studies that are cross-sectional (Ellwood and Greenwood, 2016;Hooper et al., 2005) or covering only short periods of time (West and Carnegie, 2010;Carnegie et al., 2013). In contrast, this longitudinal investigation spans 28 years and allows an understanding of how and why responses to the valuation requirement and related controversy have unfolded over such an extensive time period. It is not merely a case of updating prior findings; the key interest is on understanding how recent changes relate to prior reactions.
Second, moving from earlier literature on the topic, the study considers which (if any) of the key anticipated risks relating to the monetary valuation of heritage have manifested, especially in the areas of accountability and commercialisation of the cultural sector (see, for example, Greenwood, 2016: Abdullah et al., 2018). In doing so, this study recognises the call by Hooper et al. (2005, p. 428) to conduct research on what "might be the effectsintended and unintendedof both mandating and resisting particular forms of accountability". Although Australia, together with New Zealand, are somewhat unique around the globe in terms of compulsory recognition and reporting of these cultural heritage collections as assets for financial reporting purposes, the study can inform debates in contexts where heritage collection capitalisation is currently being considered.
Lastly, in applying the theoretical lens of paradoxical theory, this longitudinal study provides the ability to report and explain with finesse the audited financial reporting experiences of 16 purposely selected major cultural institutions in Australia. In short, the main benefit of applying paradox theory is to stimulate a more nuanced understanding of the collective institutional reactions to heritage monetary valuation requirements. Although the concept of a paradox is often used in accounting research to critically interpret the effects of changes in disclosure requirements and accountability (Bukh, 2003;O'Dwyer and Unerman, 2008), technology (Al-Htaybat and von Alberti-Alhtaybat, 2017), and the accounting and auditing profession (Gu enin-Paracini and Gendron, 2010; Annisette and Trivedi, 2013), only a few studies have specifically addressed the key responses to paradoxical tensions in a theoretically informed way (i.e. Amyar et al., 2019;Lewis et al., 2019). The present study contributes to this nascent stream by examining paradoxical responses at the field level, based on the 16 major cultural institutions, rather than at a specific organisation level. In addition, the longitudinal approach enables patterns of responses over time to be observed, hereby providing insight on how the paradox evolves as responses unfold.
The remainder of the paper is structured as follows. In the next section, a definition of a paradox is provided, and the prior literature is drawn upon to define heritage valuation as a paradoxical practice. The analytical framework applied is presented, which is informed by recent studies on the processual nature of paradox and paradox management. In the methodological section, information relating to the study's sampling, data collection and analytical procedures is provided. Four phases make up the findings section, each of them titled in accordance with the paradoxical responses that characterise them. After a synthesis of the findings, the discussion section outlines the implications for the debates on heritage financial reporting and paradox management. The concluding comments close the paper, with reference to the coronavirus disease 2019 (COVID-19) pandemic crisis and the potential The paradox of accounting for cultural heritage implications of accountants' mandatory financial valuation of heritage assets on the very existence of global cultural institutions.
2. Analytical framework 2.1 Heritage valuation as a paradoxical practice As a key contribution to the literature, this study applies paradoxical theory to a new context as a framework for the understanding and development of heritage financial reporting by public, not-for-profit cultural institutions since the early-to-mid 1990s in Australia. The word "paradox" is of Greek origin and comes from the word "par adoxon", a composite of "para" (contrary to) and "d oxa" (opinion). A paradoxical situation is one in which elements that seem logical in isolation meet, leading to outcomes that are absurd and irrational (Lewis, 2000); indeed, heritage professionals and accountants attribute value to heritage collections using their own distinctive logic. For heritage professionals, values have been described as multiple, mutable, incommensurable and often in conflict (de la Torre, 2013;Alonso Gonzales, 2016;Qassar, 2020;G omez Villa, 2018;Kompatsiaris and Chrysagis, 2020). As Smith (2006, p. 11) explains, heritage is not a collection of things, but "a process, or a performance", in which multiple stakeholders identify and negotiate aesthetical, historical, cultural, social or recreational meanings. Given the multiplicity of values, there is no common measure that can be applied, and this leads to both incommensurability and incomparability (de la Torre, 2013). Further, heritage is often politicised and contested, and its conservation is not a neutral practice (Macdonald, 2009;2015).
While heritage professionals and scholars highlight the importance of multiple systems of evaluation (Lamont, 2012) expressed by different stakeholders, in the context of financial reporting the definition of monetary value, as either value in exchange or value in use, is defined from the standpoint of the reporting entity and, therefore, is less inclusive. Value in exchange equates value with market price (Rowles, 1992) and for instance, presupposes the existence of an organisation that can set or accept a price for a work of art, no matter what value(s) other stakeholders may consider appropriate. Alternatively, value in use, represents the subjective value of an object for its owner (Stanton and Stanton, 1997) and raises the complex issue of who owns the artefacts comprising heritage collections as well as the legal status of public collections (Carnegie and Wolnizer, 1995;Stanton, 1997Stanton, , 1998. Not-for-profit, public cultural institutions are the custodians, not the owners of the collections (Carnegie and Wolnizer, 1995). These authors argue "by virtue of their dedication for community use, items in collections are off the market; they are exempt from market forces" (1995, p. 41). Collections owe their creation and existence to being set purposely aside in a special domain, where artefacts are protected by the Acts of Parliament which enabled these institutions to be established (for example, in Australia), or by Constitutions and, as a result, they are effectively removed, at least in conceptual terms, from the cut-and-thrust of the economics of the market place. Besides, contemporary museology practice recommends the restitution of illicit, sacred or ceremonial objects to their original communities (ICOM, 2016), that is, to where they legitimately belong for appropriate conservation and preservation.
Consistent with the definition of a paradox outlined above, accounting literature on heritage valuation has underlined the absurd and irrational outcomes emerging from the encounter between the two logics. At a theoretical level, Wolnizer (1995, 1999) have shown how collections do not satisfy the formal accounting definition of an asset. The authors challenge those who advocate heritage capitalisation to answer key questions, six in all (reproduced in Appendix 1). Indeed, the authors, in arguing against this financial reporting practice, describe it in unflattering terms as an "accounting fiction" (Carnegie and Wolnizer, 1995) and an "accountability mirage" (Carnegie and Wolnizer, 1996).
Interestingly, the specific questions put forth by Wolnizer (1995, 1999) remain unanswered at the time of writing. The relevance of these questions to this longitudinal study relates to the perceived lack of accountability of the accounting profession (Carnegie and West, 2005;Woon et al., 2019) in not defending or supporting this mandated practice. In plain terms, the accounting profession is pushing for, or even demanding, such prized and unique organisations to be more accountable. Paradoxically, however, the accounting profession in not answering these specific questions is, in itself, perceived as lacking in accountability for its own actions.
On empirical grounds, West and Carnegie (2010) and Carnegie et al. (2013), in the context of the library collections of public universities in Australia, investigated the valuation practices for financial reporting purposes of 36 university libraries during the periods and 2007, respectively. Echoing Miller (1998, both studies refer to heritage valuation as a "chaotic accounting margin", characterised by perturbing differences between universities and frequent inconsistencies "that have no clearly discernible rationale" (West and Carnegie, 2010, p. 223). This has paradoxical consequences with this "accounting" being questioned as to whether it indeed hinders rather than aids adequate assessments of the financial performance and position of Australia's public universities (West and Carnegie, 2010).
Other empirical studies have also shown that the financial value of heritage collections is of little relevance for users of financial statements, who are more interested in the costs of preservation or the descriptive information about the physical condition of the collections (Aversano and Christiaens, 2014;Aversano et al., 2019; also see, for example, Adam et al., 2011 andAversano et al., 2015 in the context of heritage reporting in local government). Furthermore, the reliability of the information has been criticised Wolnizer, 1995, 1996;Barton, 2000Barton, , 2005Lapsley et al., 2009;Biondi and Lapsley, 2014). According to Carnegie and Wolnizer (1995, p. 43), "markets are thin because of the unique, one-of-a-kind nature of many of the items [or artifacts]. Consequently, reliable prices for 'similar' items are not readily available, and it is difficult to achieve a reliable consensus among appraisers". Further, Biondi and Lapsley (2014. p. 146) declare the "deep seated, pernicious problems of asset recognition and valuation" for public heritage assets. In view of these circumstances and conditions, it is understandable for external auditors of cultural institutions to face the attestation of the financial valuation of collections of "priceless objects" (Leggatt, 2019, title) and "priceless treasures" (Tam, 2020, title) of "immeasurable worth" (Connolly, 2019, title) with some trepidation.
An interesting feature of the heritage valuation debate is that, since the beginning, authors have pinpointed several potential risks associated with this practice. Amongst the effects of heritage valuation, many authors (Carman et al., 1999;Carnegie and Wolnizer, 1996;Hooper et al., 2005) have highlighted the impact of government-imposed financing or "capital" charges or the negative effects of depreciation expense on museums' net profit. Further, drawing on Shr€ odinger's cat paradox, Ellwood and Greenwood (2016) claim that a second hypothetical risk embedded in the monetary valuation of public heritage is that measuring the economic value of heritage may decrease its cultural value and called for investigations on how measuring economic value affects cultural value. This may lead to deaccessioning decisions based on the need to improve an institution's cash flows, for instance, rather than based on curatorial assessment (see, for example, Cultural Heritage Agency, 2014, for a heritage professionals' guide to cultural valuation). In their study, Ellwood and Greenwood (2016) report on the case of a British local government deaccessioning a statue following its economic valuation. However, the authors also acknowledge the lack of data to infer causality between the two events.
More generally, the monetary valuation of heritage for financial reporting may be responsible for distorting, rather than improving, the accountability of museums' managers. Valuing public collections in monetary terms may overemphasise the financial assessment of objects and thereby de-emphasise other assessments of institutions The paradox of accounting for cultural heritage (Carman et al., 1999), where the latter is more important than the former to ensure a vital and viable cultural sector (Carnegie and West, 2005). Besides, for Carnegie and West (2005), the custodians of cultural heritage resources are not accountable for maintaining or maximising the financial values of heritage, rather under the missions of these institutions, they are to conserve and preserve these resources, and do not have the power to trade commercially in heritage collections. Lastly, the monetary valuation of heritage may also be a risky field for accounting professionals. On the one hand, the irrelevance of knowing the value of a collection for accountability purposes, coupled with the unreliability of such values and the high cost of obtaining this information, could emphasise mannerisms such as "unnecessary jargon or unjustifiably esoteric techniques or pseudo-paradigmatic changes" (Larson, 1977 in West andWest, 2005, p. 922). On the other hand, the "nonsense" or "bizarre results" emerging from financial valuation practices observed at the organisational level (Hooper et al., 2005;West and Carnegie, 2010;Carnegie et al., 2013) could indirectly communicate that "any value is acceptable" when accounting figures are concerned, including, for instance, subjective and arbitrary choices, hence delegitimising financial reporting, the audit opinion statements, the accounting discipline and the profession (see, for example, Parker, 1996;Barton, 1999;West, 2003, 2005;Ellwood and Newberry, 2006 commenting on public sector accounting under New Public Management or NPM).
As discussed above, ascribing a monetary value to heritage collections creates a paradoxical situation. This is because the practice, which was intended to improve decisionmaking, comparability and accountability can have the opposite effect. Rather, this practice produces, information that has scarce relevance to users, if any at all (Aversano and Christiaens, 2014), conceptually misplaces the "measured" collections, reduces the ability to draw meaningful comparisons, and has the potential to jeopardise accountability. Such arguments have been repeatedly made at least from the early to mid-1990s once this form of accounting under NPM appeared on the agendas of public cultural institutions.

Process perspectives on paradox management
Organisational research on paradox management has shown that paradoxes and the organisational responses to them evolve over time (Smith and Lewis, 2011;Jarzabkowski et al., 2013;Jay, 2013;Hargrave and Van de Ven, 2017;Raisch et al., 2018;Smith and Besharov, 2019). This is important to acknowledge in setting out to identify and analyse the practice of financial reporting of cultural heritage in Australia over a continuous period of 28 years. Although differences between processual models of paradox management exist, the frameworks used are similar in that they each involve analysing the nature of a paradox. Further, they each recognise the reactions to the paradox, and the outcomes that occur, as well as the influence of actions on the contradiction itself (Hargrave and Van de Ven, 2017). Such actions may be continuous across time or variable as illustrated in this study.
In terms of the nature of a paradox, organisational paradoxes may evolve from being latent, that is, unnoticed or dormant, to salient, meaning actively requiring a response. The nature of a paradox shifts from latent to salient because of conditions which emphasise the functional interdependence between the opposing logics involved. These include pluralitythat is, the presence of a multiplicity of views making conflicting goals emergerapid change stimulating quick alternative selection and scarcity of resources that poses "either that or" dilemmas (Smith and Lewis, 2011;Hargrave and Van de Ven, 2017). In addition, Jay (2013) explains that paradoxes become salient when organisational actions can be interpreted either as a success or a failure depending on the logic employed to make sense of them.
When a paradox becomes salient, it motivates a response. Strategies to cope with paradoxes can be categorised as either defensive or active (Lewis, 2000;Jarzabkowski et al., 2013). The former category of strategies are regarded as "paralysing defences, which initially reduce discomfort and anxiety, yet eventually intensify tensions" (Lewis, 2000, p. 762). Key forms of defensive responses include the following: "splitting", which involves dealing with one pole of the paradox at a time or compartmentalising poles in distinct spaces or hierarchical units; "selecting" or "suppressing", which involves prioritising one element and allowing it to dominate or overrule the other element of the paradox, and "opposing", that "involves parties supporting contradictory elements of a paradox engaging in active confrontation and conflict that polarise paradoxical elements" (Jarzabkowski et al., 2013, p. 255).
Active responses, on the other hand, try to engage with the competing forces embedded in a paradox and can enable change (Lewis, 2000;Jarzabkowski et al., 2013). Active strategies are "acceptance", which is learning to live with a paradox, "adjusting" or "synergy", that involves finding a way to respond to both sides of the paradox along with "transcendence", meaning reframing the tension, such that the two poles are no longer regarded as opposite.
While for Smith and Lewis (2011), the choice between active and defensive reactions is affected by the cognitive features of actors and the organisational dynamic capabilities, Hargrave and Van de Ven (2017) highlight the role of power distribution to explain the variance amongst responses. For these authors, under a condition of symmetrical power distribution, an opposing reaction is unlikely since it will most probably be ineffective for the coalitions backing each pole of the paradox. Hence, actors will either engage in the active strategies described above or employ a "mutual adjustment" which means accepting the coexistence of the two positions and, thereby, producing "mutually satisfactory but not necessarily mutually advantageous outcomes" (Hargrave and Van de Ven, 2017, p. 331). Conversely, when power is distributed asymmetrically, those with less power will seek to oppose the paradox, with a view to stimulating change. Those with more power, however, will try to incorporate practices and arrangements which have been associated with the subordinate element aspiring to "assimilation". Rather than being a single occurrence, these responses to paradoxical tensions cumulate over time affecting whether the paradox is reproduced, revised or transformed (Jarzabkowski et al., 2013;Hargrave and Van de Ven, 2017;Smith and Besharov, 2019). Jarzabkowski et al. (2013) explain that responses may unfold along a defensive or active path depending on whether reactions are embedded into organisational routines. Theorising from an empirical case about organisational restructuring, these authors observed that an initial defensive response that is embedded in organisational practices can lead to further defensive reactions, eventually leading to an escalation of conflict and the perpetuation of vicious cycles. Alternatively, actors may move to an active approach when a defensive response does not become part of the procedures through which resources, attention and roles are distributed. Smith and Besharov (2019) discuss the crucial role of learning in understanding the cyclical nature of paradoxes. These two authors note that leaders learn to manage tensions between dual missions more effectively and efficiently over time, thereby making the "swings" between the two poles of the paradox both shorter and less extreme. Lastly, Hargrave and Van de Ven (2017) stress the role of unintended consequences to explain why reaction strategies to paradoxes change over time.
These analytical categories of paradox theory are adopted in this study to understand why particular outcomes relating to the financial reporting of heritage collections, as assets, in Australian public cultural institutions have occurred across time.

Context and period of analysis
The study's empirical analysis of the responses to the heritage assets valuation requirement in Australia during the period from 1992 to 2019 is based on the latest annual reports of these  (Rowles, 1992). Potter (2002, p. 75) indicates that "this discussion paper represented the first formal steps taken by accounting regulators in Australia toward developing detailed accounting regulations for non-profit cultural organisations in Australia". Indeed, accounting pronouncements, in the form of Australian Accounting Standards (AAS) and Statements of Accounting Concepts (SAC), were issued in the country between 1992 and 1998. Such pronouncements requiredimplicitly or explicitlythe recognition of heritage objects as assets in the statements of financial position of cultural institutions and other public entities (SAC 4 Definition and Recognition of the Elements of Financial Statements in 1992; AAS 29 Financial Reporting by Government Departments in 1993; AAS 27 on Financial Reporting by Local Government in 1996; AAS 31 Financial Reporting by Governments in 1998). Together with similar pronouncements issued in New Zealand on the topic, they represented an early move and, therefore, an exception in the global scenario at that time.

Method and sample
A cross-sectional longitudinal study design (Carnegie and Napier, 1996) is used to review the heritage monetary valuation practices of a purposely selected sample of major Australian public cultural institutions within the study period. The sampling procedure commenced by listing all Australian not-for-profit, public institutions preserving, cultural, heritage and scientific collections in 2019. The listed institutions fall within the jurisdiction of the Commonwealth of Australia (CoA) or the Federal Government, state governments and the Northern Territory Government. Privately owned collections were excluded since they are not affected by the accounting standards identified above and for which the paradoxical tensions are less likely to exist given the investment of private capital by personal choice. Local-level museums and art galleries were also excluded, for which, as suggested by Hooper et al. (2005), the pressure to comply with financial valuation requirements is weaker due to their relative independence from central government in terms of funding. At this point, the list was made up of 25 public cultural institutions, including state art galleries focussing on visual art (5); state museums preserving and displaying natural specimens and objects that are relevant to understand the history and culture of the state (5), combined state museums and art galleries (2); a federal level art gallery (1); a federal level museum (1) and other institutions with distinctive thematic collections (11), for example, the Australian War Memorial, the Museum of Applied Arts and Sciences, the Maritime Museum, the Museum of Australian Democracy and the Arts Centre Melbourne.
A decision was made to include in the final sample two major specialist institutions belonging to the "other institutions" category. These organisations are the Australian War Memorial, situated in Canberra, and the Museum of Applied Arts and Sciences in Sydney, New South Wales. The size and significance of their collections make them comparable to state museums or state art galleries in terms of the number of their collection items, the reported financial values of the collections and the reported financial values of collection assets as a proportion of reported total assets.
The 16 major Australian cultural institutions selected, drawn from across the six states and the two major territories of Australia, are all prominent organisations preserving and conserving unique and prized collections for community safekeeping, appreciation and understanding, and for educative purposes at the local, national and international levels. For instance, the Australian Museum, formed in Sydney in 1827, is Australia's first public museum and has the largest collection in terms of total objects held in the country [2]. In the State of Victoria, Museums Victoria was founded in Melbourne in 1854 in the newly constituted Colony of Victoria [3] and The National Gallery of Victoria, founded in 1861 in Melbourne, is Australia's oldest public art gallery (Strecker, 2002).
In summary, the study's sample of major cultural institutions includes six museums, six art galleries, two combined museums and art galleries, and two institutions with specialised collections (identified above). Each of the 16 organisations included in the sample are statutory entities receiving monetary support from their home state or territory and are formally accountable to the state or territory-level Ministry of the Arts. Table 1 provides key information (including the name, jurisdiction, number and genre of objects) for each of the institutions comprising the sample.
3.3 Sources, data collection and qualitative content analysis The annual reports published by the 16 Australian cultural institutions for the financial years 1991/1992 to 2018/2019 (that is, the last available annual report at the time of writing) are the main sources of data for the study. In some cases, we can provide evidence also from 2019/ 2020 reports, but not all the 2019/2020 annual reports were available at the time of completion of this study. Annual reports from approximately the 2000/2001 financial year onwards were retrieved from the institutions' web sites in a digital format. Printed annual reports from earlier periods were collected from the State Library of Victoria, located in Melbourne. Annual reports are particularly valuable to analyse the encounter between heritage professionals and accounting-based approaches to the monetary valuation of heritage across time. In fact, the annual reports of each institution comprise an annual review, where the achievements of the year are presented by means of narratives and quantitative financial and non-financial indicators; an independent auditor's report and the financial reports. Most of the authors of the present paper are members of the accounting profession and are proficient in financial reporting, annual reporting disclosure practices, and accounting policies and how they are applied Qualitative content analysis (Mayring, 2004; also see various contributions in Hoque et al., 2017, p. 524) was adopted and performed on the annual reports at four interrelated stages. First, the following data originating in the statements of financial position and the accompanying notes were recorded in an electronic spreadsheet. A separate sheet for each institution was used with a dedicated column for each year. Information reported in the rows referred to the carrying value of the collection; basis of carrying value; depreciation policies (see West and Carnegie, 2010 for the use of a similar approach). Any additional notes regarding the financial valuation of heritage were also recorded. Rows were added for any cost items relating to heritage valuation (depreciation; capital charge; gains or losses on revaluation).
Next, an exploration of each annual report, using keywords and key phrase searches was undertaken to assist in determining whether and how the monetary value of the collection was included in the presentation of the annual results and to identify the basis of deaccessioning decisions in the annual review. The audit opinion accompanying each financial report was also reviewed to identify any relevant audit opinion qualifications or other comments. A row for any additional source was added in each institution's electronic sheet, reporting relevant sentences word-by-word, when present. Additionally, triangulation took place involving the data collected, relevant newspaper stories or articles, and other documents collected by running ad hoc searches on the websites of the Australian Accounting Standard Board (AASB), the IPSASB, the Federal Auditor-General and State government Auditors-General. This allowed for the collection of additional relevant documentation (i.e. PAC, 2002;PAEC, 2002;New South Wales Auditor-General, 2010, CAMD, 2017, 2018. A synthesis of this evidence was also included, adding new rows in the electronic spreadsheet. Second, using the information organised into the electronic spreadsheet, reconstruction was made of the evolution of accounting for heritage financial valuation practices for each institution. A line graph illustrating the movement in the collection values over the period of 28 years for each organisation was constructed. Further, a short narrative was prepared to identify the elements relating to changes in the value of collection and accounting policies, as well as evidence extracted from the annual review, the auditor statements and the additional relevant documentation. Third, due to the understanding achieved through the preparation of institution narratives, theoretically informed coding employing the main constructs of the processual perspectives on paradox management was performed on the electronic spreadsheet. More specifically, for each organisation and each year under analysis, coding was conducted relating to whether the paradox was latent or salient, and we made a note of the discernible responses to it. The paradox of heritage monetary valuation was considered to be salient whenever notes to the financial reports accounted for a process of revision to the valuation procedures, when difficulties in such processes were highlighted and/or when changes in the collection value were justified in the annual review or, less often, in newspaper articles. Reliance on the response types available in the prior literature occurred to make sense of the reactions to paradoxical tensions. Eventually, the following key reactions by different institutions and at different points in time were observed: "splitting", "opposing", "assimilation", "suppression" and "mutual adjustment". For illustrative purposes, Appendix 2 provides examples of each of the first forms of responses, while the "mutual adjustment" designated response is clearly evident in the work of the Council of Australasian Museums Directors (CAMD or Council) as further discussed later.
Thereafter, the identification of meaningful ways to interpret the heterogeneity that was emerging at the organisational field level was sought. Creating clusters was not effective, as it was difficult to pinpoint groups of institutions sharing the same responses along the consecutive period of 28 years. Eventually, the dynamic of change over time was examined more specifically allowing the pinpointing of four distinct phases, each characterised by a distinctive nature of the paradox (latent vs salient) and a totally or partially new combination of responses. For some cultural institutions, the differences between the four phases are very marked, while for others, the responses to the paradox remain more stable over time. However, although not all movements or actions fit neatly within the identified phases, the overall trends can be clearly articulated. After having pinpointed the four phases, attention was focussed on how the outcomes of each phase triggered reactions in the next one, consistent with the notion of a paradox-response cycle (Jarzabkowski et al., 2013). The four phases cover the following unequal financial periods : 1991/1992-2001/2002, 2002/2003-2009/ 2010, 2010/2011-2015/2016 and 2016/2017-2018/2019. 4. Findings: responses to heritage valuation amongst Australian cultural institutions 4.1 First evolutionary phase: splitting and opposing (1991/1992-2001/2002) 4.1.1 Nature of the paradox and reactions. At the beginning of the 1990s, the issue of valuing heritage for financial reporting became salient for most, if not all, cultural institutions included in the sample. Triggered by changes in the accounting pronouncements, institutions began to contemplate the monetary valuation of their collections, although with different perspectives and approaches and at a different pace. Some institutions decided to start the process by reporting only partial valuations. This is the case for Museum Victoria, the National Gallery of Victoria, the Art Gallery of Western Australia and the Queensland Art Gallery, which were each partially valuing their collections even before 1991/1992. The National Gallery of Australia started to do so in 1991/1992, the National Museum of Australia The paradox of accounting for cultural heritage commenced in 1994/1995, the Art Gallery of South Australia and the South Australia Museum each started from 1996/1997. As an example of partial valuation, the National Gallery of Australia from 1991/1992 to 1996/1997 recognised as an asset only works of art acquired by way of gift from 1986/1987. These assets were valued in monetary terms at cost based on the curators' valuation or, alternatively, at an average of a series of expert valuations, which were indeed not priceless. The partial assessment of the collections evolved to comprehensive valuations by 2001/ 2002, resulting in huge increases in the financial values reported for their collections and a corresponding lift in the reported values of the institutions' total assets. The approach of some institutions such as the Australian War Memorial, the Museum of Applied Arts and Sciences, and the Queensland Museum in this period was only slightly different. These organisations took a longer period to initially place financial values on their collections but commenced doing so by reporting total, rather than by firstly adopting partial, monetary valuations. Indeed, until 1998/1999 the Museum of Applied Arts and Sciences valued its collection "at the nominal amount of $1 as it is considered that values for the items in the collection cannot be measured reliably" (MAAS, 1999, p. 32). Lastly, and somewhat exceptionally, the financial value attributed to the collection of the Art Gallery of New South Wales pertained to the whole collection from as early as 1991/1992.
While the cultural institutions in the sample progressively implemented the regime of identifying monetary valuations for financial reporting purposes, interestingly, no mention of the newly developed accounting figures can be observed in the annual reviews, where the yearly results and performance indicators of the organisations are presented. In other words, what was required to be done in accounting terms was not seen as worthy of mention in reporting on the operations the institutions. In this phase, no asset-based ratios were presented alongside other measures of performance, which included achievements along the curatorial, visitor and resource dimensions (Zan, 2006) [4]. The confinement of the asset values for collections to the financial statements themselves suggests that the introduction of a new, alternative valuing logic was, in many cases, managed through a "splitting" response (Lewis, 2000;Jarzabkowski et al., 2013), which involves separating contradicting elements of a paradox into different realms. There were, however, two notable exceptions.
In a somewhat radical fashion, the Australian Museum and the Western Australia Museum contested the placement of monetary values on their collections. The Western Australia Museum justified the non-recognition decision by referring to "the intrinsic value of such acquisition and in accordance with policies generally followed by similar museums in North America" (WAM, 1996, p. 73). In addition to reliability issues, the Australian Museum emphasised the dissimilarities between the heritage professionals and accounting perspectives on value. For the oldest not-for-profit public museum in Australia, it was argued that the collections' "true value cannot be expressed in monetary terms but rather in terms of their cultural and scientific worth" (AM, 1999, p. 14). Both responses anticipated Barton's (2005) argument on the need to treat heritage as public assets, thus removing them from the market and disregarding the associated logic to account for them in financial reports.
Amongst responses to paradoxical tensions, "opposing" involves active confrontation between parties and further polarisation of positions (Lewis, 2000;Jarzabkowski et al., 2013). In fact, the resistance of the Australian Museum led the Auditor General of New South Wales to controversially qualify the museum's accounts for non-recognition of collection assets in their statements of financial position in 1998/1999. Following the first qualified opinion, the museum commenced the valuation process that led to an initial estimation of monetary value of $4bn. However, the Museum itself did not consider this attempt at valuation to be reliable and, therefore, did not include the figure amongst their non-current assets but rather disclosed this figure in the notes to the financial statements. Therefore, the New South Wales Auditor-General qualified the audit opinion in the Museum's accounts again in 1999Museum's accounts again in /2000Museum's accounts again in , 2000Museum's accounts again in /2001Museum's accounts again in and 2001Museum's accounts again in /2002. Triggered by these events, the New South Wales Public Accounts Committee inspected the Australian Museum in November 2002 and put additional pressure on the museum to begin capitalising its collection and recognising this asset in its statement of financial position. This eventually happened in 2002/2003, when the Australian Museum recognised, for the first time, a collection value of $530.1m [5].
In the same period, the Parliament of Victoria Public Accounts and Estimates Committee's (PAEC) critical observations on the valuation requirement did not receive much attention. The report presented the critical opinions of cultural institutions on the costs, benefits claimed to be involved and the reliability of valuing heritage in monetary terms (PAEC, 2002). Building on this evidence, PAEC formulated several recommendations to the Victorian Government Department of Treasury and Finance, including to request to the AASB for a reexamination of the requirement for public sector agencies to ascribe monetary values to collections, with consideration given to the non-recognition approach adopted in many other major countries.
While appreciating the call for international consistency, the government did not support the recommendation on the rather limp basis that "it might be seen as support for those seeking not to value collection or heritage assets" (Government's Response, 2003 p. 3). Why putting a monetary value on collections was deemed more important than international harmonisation was not further specified in the document. The unsuccessful opposition of PAEC and the end of the "opposing" response by the Australian Museum marks the end of the first phase in 2001/2002. The Australian Museum was one of the two institutions which mounted "open opposition" to collections monetary valuation. As the Australian Museum was the most vocal voice against financial valuation, the decision was made to close phase 1 when this high profile and significant opposition ceased.
4.1.2 Outcomes and link to the next phase. Following the introduction of new accounting pronouncements in phase 1, institutions whose executive/board management were not valuing collections in monetary terms started to do so, albeit reluctantly given the perceived lack of benefits of this commonly perceived "new accounting technique" within the institutions. Partial valuations were increasingly completed, yet through many different valuation approaches. Overall, the astonishing increase in the total financial value of collections in this period (from $539m in 1991/1992 to almost $6bn in 2001/2002) is explained by the progressive convergence towards monetary valuation rather than to actual increases ascribed to the physical collections.
The progression of monetary valuation proceeded in tandem with the "splitting" practices of most institutions and the silencing of the critical stances. The weakness of the opposition (involving a few museums and a Victorian initiative) and the rigidity of those to whom the critiques were directed will now be shown to magnify the contradictions related to heritage valuation that will be observed in the second phase.
4.2 Second evolutionary phase: further splitting and assimilation (2002/2003-2009/2010) 4.2.1 Nature of the paradox and reactions. During this phase, tensions between the accounting and heritage professionals' approaches to valuation became latent. Distinct from the prior phase, and the next phases, in this phase there is no evidence of discussion around the financial valuations of the collections in any of the annual reviews or auditor statements analysed. In addition, no external documentation was collected on this issue. This is the result of a combination of two responses. Firstly, cultural institutions continued to employ the "splitting" approach observed in the prior phase, whereby heritage is valued in a perfunctory way. In these instances, the figure is not "talked about" and has no apparent effects on actions (such as any deaccessioning decisions). Secondly, in this phase manifestations of an "assimilation" response (Hargrave and Van de Ven, 2017) were observed.
The paradox of accounting for cultural heritage "Assimilation" occurs when "practices and arrangements which have been associated with a subordinate element come to be incorporated into the dominant element" (Hargrave and Van de Ven, 2017, p. 331). By accepting selected elements of the opposite logic, the dominant party maintains its position and avoids active confrontation. In the authors' assessment, auditors' tolerance of the extreme diversity in accounting practice in a firstmover nation reflects an "assimilation" response in this context. Even though cultural institutions appear to be using the same method for valuing their collections (for example, fair value however interpreted across institutions), there were significant differences observed in the identified methodology and techniques adopted from 2002/2003 onwards, with adverse effects on comparability and even on understandability. Such differences included the frequency of revaluations every three or five years or annually using an indexation approach such as at the Tasmanian Museum and Art Gallery and the Australian Museum, and whether sampling techniques are employed, as at the National Gallery of Australia, the Art Gallery of Western Australia and the Art Gallery of South Australia.
Tolerance towards blanket exclusions of particular asset types is a further example of "assimilation". For example, the Queensland Museum stated, until 2007Museum stated, until /2008, that "culturally sensitive, secret and sacred objects and human remains were not valued or verified" (QM, 2008, p. 88 [6]. Allowing the exclusion of secret and sacred objects for reasons of cultural sensitivity demonstrates assimilation by accepting a museological deterrent to monetary valuation into the accounting logic, conceding, in an implicit way, in the authors' view, the non-reliability and non-relevance reasons for non-recognition. At the same time, considering only Aboriginal peoples' secret and sacred objects as culturally sensitive, which is completely understandable, undermines the heritage professionals approach of emphasising multiple sensibilities in any object, such as objects extracted without any consent from dominant regimes or empires from suppressed colonised or invasion ravished communities, territories and nations. Lastly, an "assimilation" response seems to have been adopted in relation to accounting depreciation policies. While most institutions choose not to depreciate their collections for financial reporting purposes, three Federal government-funded institutions, all located in the Canberra (National Gallery of Australia, Australia War Memorial and National Museum of Australia), charged depreciation on the heritage assets recognised in their statements of financial position, using the straight-line method and with frequent changes made in the useful lives adopted. For instance, the National Museum of Australia began depreciating its collection for accounting purposes in 2000/2001 and, until 2002/2003, used a useful life ranging from 75 years to unlimited. The range was changed to between 50 and 500 years in 2003/2004. Eventually, the maximum useful life was increased to 5,000 years starting in 2004/2005. In the authors' perception, auditors' tolerance for a 5,000 years depreciation policy affirms the strong museological claim that collections do not lose value in cultural terms because of preservation and conservation endeavours while, at the same time, allowing the auditor and institutional management to avoid active confrontation. Why collection artefacts are assigned a 300-years or 5,000-years useful life is indeed not disclosed or discussed in the institutions' annual reports. In the authors' view, this level of tolerance would be most unlikely to be witnessed in any other organisational context, which itself is a major point of differentiation of not-for-profit, public cultural institutions.
Do such extremities of useful lives reflect different conservation policies? Does it suggest that the 5,000-year artefact is more culturally relevant than one estimated and set at 300 years? Discussing depreciation policies amongst Australian public universities' libraries, West and Carnegie (2010) state that rather than reflecting distinct collection management strategies, differences between depreciation policies are the consequence of a regulatory void. As they point out, "when there is no consensus or regulatory edicts specifying what accounting is 'right'', it is presumably difficult to establish that any particular approach is 'wrong'" (West and Carnegie, 2010, p. 220). It is believed that a similar comment can be extended to the depreciation policies amongst the Federal government-funded institutions during this phase.
4.2.2 Outcomes and link to the next phase. From a quantitative point of view, the pace of revaluation slows down in phase 2 (reported collection values change by an average of more than 11% per annum vs just over 32% during phase 1). More notably, the joint effects of "splitting" and "assimilation" responses make the tensions become latent and unnoticed. However, at a deeper level, "contradictions" (Hooper et al., 2005) "bizarre results" or "nonsenses" (Carnegie et al., 2013) emerge, having an adverse effect on the comparability of the information produced by means of the financial valuation of the collections. The contradictions due to the arbitrary nature of attempting to assign financial values to assets that are essentially held for non-financial, social purposes will re-surface in the next phase. This phase 3 development is related to the number of organisations dealing with heavy, lump sum write-offs in collection values, bearing consequences in the ways in which their financial performance was publicly represented. The write-down season commenced at the Tasmanian Museum and Art Gallery in 2010/ 2011 when the museum engaged a new valuer to undertake a revaluation of their collection. After introducing some changes to the valuation methodologies and techniques employed, such as valuing the numismatic collection using fair market value rather than on the weight of precious metals at current market rates, the valuer's activities led to a write-down of $116m, representing close to a quarter of the 2010/2011 opening value of $487m.
In the same year, the Art Gallery of South Australia reported a write-off of just over $38m, which represented approximately 7% of the collections opening value that year. This was the result of the introduction of a more conservative policy regarding how to reconcile disagreements between the valuations undertaken by internal curators and external valuers. Between 2002Between /2003Between and 2010Between /2011, the valuation policy of this institution favoured the higher valuation obtained for artefacts, whether it was provided by the internal or external valuer. However, by 2010/2011 the institution had adopted a new policy, according to which "in the event of variations between the values of the internal curator and the external valuer, [the gallery] is to adopt the average value" (AGSA, 2011, p. 91). This new valuation policy was not explained or justified, appearing to be an indication of the "any figure is okay" mindset as an average value is not verifiable by any evidence in itself. A smaller write-down due to changes in valuation methodologies was also registered at the National Gallery of Victoria in 2015/ The paradox of accounting for cultural heritage 2016, when the value of the collection was reduced by 6% due to the employment of a new stratified multistage sampling technique. Differently from the cases described above, write-downs at the Museum of Applied Arts and Sciences (2014/2015) and South Australia Museum (2015Museum ( /2016 relate to the exclusion of specific parts of the collections from valuation. The Museum of Applied Arts and Sciences reported that the $95m decrement related to the Boulton and Watt steam engine, which is the oldest rotative steam engine in existence and one of only a few remaining in the world which remains operational. The asset was excluded on the basis that "the revaluation identified the Museum's 1785 Boulton and Watt steam engine as being unique and not able to be reliably measured" (MAAS, 2015, p. 41). The write-down represented almost a quarter of the 2014/ 2015 opening value of "heritage assets" amounting to $403m. What is of particular note in this instance, different from the other cases, is that the Museum of Applied Arts and Sciences' write-off was only partially absorbed by the revaluation reserve, thus causing a loss on revaluation in the income statement, and eventually leading to an operating loss being recorded for the financial year of $67,869. Most of the write-down recorded at the South Australia Museum (equal to 20% of the opening value) is due to the exclusion of the foreign archaeology collection from the 2015/2016 valuation for reasons that are not explained in the annual report.
The write-down registered by the Australian Museum in 2012/2013 ($375m or almost 44% of the 2011/2012 collection valuation) is the largest amount observed during the period, raising serious doubts on the reliability of the financial information reported in the statement of financial position by the Museum in prior years. In 2012/2013, the Museum acknowledged that prior valuations included "unregistered collections [that] cannot be reliably valued as they are not yet identified and recorded" (AM, 2013, p. 45). These collections were valued in monetary terms on an estimate of the size of potential collection material. In explaining how the value of the collection came to be inflated by hundreds of millions of dollars, the museum's executive officer confessed that between 2003 and 2011 "a period of increased field activity coincided with the practice of including estimates of unregistered, unsorted items in the valuation process" (Neill, 2014, p. 3). Issues with cataloguing procedures were also identified during audits performed by the Auditor-General in 2010 and 2011 (New South Wales Auditor-General, 2010;. Indeed, when commenting on the massive write-down in an interview to The Australian newspaper, the incoming Director of the Australian Museum played down the revaluation significance stating: "We've got 18 million items in the collection . . . there are still 18 million items-plus in the collection. There is a valuation methodology that was changed ... the collection has not changed" (Kim McKay quoted in Neill, 2014, p. 3). In other words, it is the non-financial value of these objects that is important and not their purported financial values, or overall movements in such values.
This McKay quotation is telling under paradox theory for at least two reasons. Firstly, it demonstrates the failure of the "splitting" response employed so far: a 44% reduction in the collection value could not be ignored as it could raise doubts over the apparent loss of the cultural, heritage and scientific values of the collection. Secondly, it shows the emergence of a short-term, defensive response: that is, "suppression" (Jarzabkowski et al., 2013). Indeed, by minimising the relevance of the information regarding the financial value of collections, the Australian Museum management was effectively prioritising the heritage professional view on value over the accounting one [8].
Another financial valuation debacle took place at the Western Australian Museum, which then triggered responses similar to those observed at the Australian Museum. After undertaking an upward revaluation in 2008/2009 which effectively tripled the reported value of the collection to $630m, the value of the collection was written down to just $347m in 2013/ 2014. This reduction was the result of a revaluation decrement of $92m and the correction of prior year errors of $199m, which were stated to occur in the earlier 2009 valuation. In similar events to those at the Australian Museum, the Director of the Western Australian Museum dismissed the financial write-down as a just a change in the "paper value" of the collection, bearing no relationship with actual changes in its size or quality (WAM, 2014, p. 9). However, the apparently artificial increase of the collection value in 2009 had very practical implications. According to the major state newspaper, The West Australian, the Museum overpaid insurance premiums from 2009 to 2011 due to this "accounting mistake" (Bevis, 2014). This is reflected in the income statements, where insurance premiums soared from $186,000 in 2008 to $309,000 in 2009, thereby reducing the institutions' operating result as a result of apparently fallacious financial values of collections [9].
4.3.2 Outcomes and link to the next phase. Owing to these write-offs, the total value of the collections increased by an average of barely 3% per annum in this period. As write-offs in these collections' monetary values started affecting cultural institutions in Australia, tensions relating to the heritage valuation practices for financial reporting purposes resurfaced. "Splitting" responses could not always be maintained because of negative publicity and reductions in operating results, implying at least deteriorating financial performance and position, as reported, with ramifications for predicting future results.
While in the immediate aftermath of the write-downs, two of the most prominent Australian museums reacted by publicly dismissing the accounting representation (Bevis, 2014;Neill, 2014). These phase 3 write-down episodes contributed to stimulating a collective and deeper discussion and debate over heritage financial valuation. This will be discussed next in phase 4. In late 2016, the CAMD, which comprises the directors of all institutions included in the sample, except for the six, specific purpose public art galleries, began drafting a framework for the valuation of collections. The CAMD was assisted by accounting professionals, and by members of the AASB and the IPSASB. The initiative originated from discussions regarding the "shared challenges that museum[s] face in measuring the value of collection[s]" and aimed at "establishing consistency and a high level of comparability of financial reporting across the Australian Museum sector" (CAMD, 2018, p. 7). One year later, in 2017, the council was the only heritage professional body to reply to the IPSASB's CP (IPSASB, 2017), mentioning the framework in its submission, which was under development at the time, as an attempt to develop cultural heritage categories to help inform financial valuation (CAMD, 2017).
Released in November 2018 "for use by the Australian Museum sector" the national guidance framework separated collections into five classes with distinctive implications for valuation (CAMD, 2018, title page). More specifically, under the framework, highly significant and sensitive collection assets are to be recorded at a nil (monetary) value; high value collection assets are to be recorded at their market value; and general collection assets can be valued using recollection cost. Further, the framework suggests employing a random sampling technique to value uncatalogued and organised objects. In contrast, uncatalogued and unorganised objects should be excluded from valuation. Guidelines on sampling techniques, valuation cycles and transparency considerations complete the framework.
Distinct from the "suppression" response adopted by a few museums in phase 3, the process through which the framework was developed seems to reflect a shift towards a "mutual adjustment" approach (Hargrave and Van de Ven, 2017) involving a coalition of cultural institutions. "Mutual adjustment" implies accepting coexistence between different and conflicting logics, yet without looking for mutually advantageous arrangementsthat is, "synergy" (Hargrave and Van de Ven, 2017). Besides, the response is possible when power is The paradox of accounting for cultural heritage symmetrically distributed. In fact, while in prior phases attempts at changing or criticising the status quo were raised by single institutions (recall phases 1 and 3), the recent framework originates from a joint response by all of the museums included in the study sample. Such a collective approach to the issue of heritage valuation is a means of re-balancing the power relationship between cultural institutions and the proponents of the monetary valuation (i.e. professional accounting bodies, accounting firms, Auditors-General and accounting standard setters). The framework does not challenge monetary heritage valuation (it accepts the notion of coexistence as a starting point) but, at the same time, it does not address any potential implications of valuing, in financial terms, the collections for asserted accountability purposes. Indeed, Australian museums' directors, in the Council's CP submission (CAMD, 2018), still disagree with the view that "suggests that valuation represents a benefit to users" because, amongst other things, the claim in "the [IPSASB] proposal that monetary values increase the inherent understanding of a heritage asset for management purposes is demonstrably false" (CAMD, 2017, pp. 3-4). Hence, by means of "mutual adjustment", the parties involved have done little, or next to nothing, but try to seek to establish a common code of conduct.
Whether any material change in approach to this controversial issue arises under the CAMD framework remains to be observed and assessed. For now, the financial statements issued after the release of the CAMD framework provide mixed evidence. Of the four institutions that revalued their collection in 2018/2019 or 2019/2020, only the Western Australian Museum reported that the valuer used the CAMD valuation framework. In line with a "mutual adjustment" approach, the Western Australian Museum underlined that the valuation was conducted exclusively to comply with general purpose financial reporting standards because "cultural assets belonging to public institutions are not intended to be sold, due primarily to legal, financial and ethical restrictions placed upon their use and/or disposal" (WAM, 2009, p. 19).
It is difficult to ascertain whether the revaluation performed at the Australian War Memorial in 2018/2019 has been informed by the framework, although the rationale of using different techniques for different classes of items (that is, full valuation for iconic items and sampling for homogenous populations of items) and the decision to no longer depreciate heritage assets are consistent with CAMD guidelines. On the other hand, the valuation technique employed at the National Museum of Australia in 2018/2019 does not differ from the 2015/2016 one, with the Museum also unusually continuing to depreciate its collection on a straight-line basis. Notably, in both cases the 2018/2019 revaluation resulted in a writedown, yet with smaller decreases compared to those witnessed in phase 3 (equal to 2 and 4%, respectively) [10]. Amongst the institutions which did not take part in the CAMD initiative, only the National Gallery of Australia revalued its collection, yet providing very little details on the approach adopted in 2019/2020.

Discussion
This study maps how a purposely selected sample of major Australian not-for-profit cultural institutions reacted to the paradoxical tensions originating from the encounter between a heritage professional and an accounting approach to the valuation of collections. Four phases emerged from the cross-sectional longitudinal analysis undertaken. Each is characterised by a distinctive nature of the paradox (latent or salient), the specific responses to it and the resulting outcomes.
In phase 1, most of the cultural institutions responded to the introduction of the heritage valuation requirement by capitalising the value of the collection (partially or totally) but without referring to these financial values in the presentation of the year's achievements ("splitting" response). Only two museums responded by "opposing" the requirement, before joining ranks with the disclosing organisations in or around 2002. The silencing of opposition led into the second phase, characterised by further "splitting" and "assimilation". The combined effect of these responses made the paradox become latent, yet under the surface, contradictions, inconsistencies, bizarre treatments and the related results were beginning to flourish. These inconsistences became evident in phase 3, following the write-downs reported by seven institutions, which drew public attention. While initially these organisations reacted to the write-downs in a short-term, defensive fashionthat is, "suppression"starting from late 2016, we observed the emergence of a collective initiative, aimed at drafting a common framework to heritage valuation for financial reporting purposes (phase 4). This initiative was interpreted as a "mutual adjustment" response. While prior studies have reported only resistance or compliance as possible institutional reactions to heritage valuation (Hooper et al., 2005;West and Carnegie, 2010), observation over an extended time period adopting the lenses of paradox theory has allowed a more diverse and complete set of responses to emerge and, thereby, be studied.
The study's results portray a roller-coaster journey for monetary heritage valuation, with variations in valuations emerging across time, as shown. In colloquial terms, these findings are not considered "a good look" for regulated financial reporting. Yet the evidence collected in this study contradicts some of the (negative) expectations regarding the consequences of financial valuation as had been put forward in prior literature on the topic. While our analysis of financial valuation practices over an extended period of 28 years does not support arguments that valuing heritage for financial reporting purposes may lead to commercial decisions and deaccessioning, this remains a risk internationally, which appears to be increasing. However, in the Australian context, deaccessioning decisions are informed by the professionally relevant repatriation policy (ICOM, 2016). In addition, these institutions have maintained missions that do not focus on maximising the financial valuation of the institutions' assets or return on assets, including the monetary values assigned to their prized collections in cultural, heritage and scientific terms.
Further, the analysis of the financial and non-financial indicators reported in the annual reviews revealed little or no apparent adverse effects of financial valuation in terms of modifying the accountability of museum management, as suggested by Carnegie and West (2005), this is, 15 years ago to the time of writing. Overall, the pragmatic approach taken by the cultural institutions analysed up to 2019 has been quite successful, by mainly discreet means, in softening or diminishing the foreshadowed unintended or negative impacts of heritage valuation. This was possible owing to "splitting" and, more recently, to "mutual adjustment" in order to minimise future "splitting" failures. Similarly, there is no clear evidence that the financial valuation of collections has limited the emergence of broader forms of reporting, as claimed recently by Woon et al. (2019). Indeed, the institutions analysed use narrative accounts and employ indicators to present their achievements concerning heritage identification, presentation, protection, conservation and transmission in line with their missions.
With this elucidation, Woon et al. (2019, p. 622)'s call for the AASB to "consider sectorspecific valuation options for such assets", while potentially commendable, remains conceptually flawed in the unique organisational and social context in which not-for-profit, public cultural institutions operate. Further, a warning is sounded against heritage accounting initiatives that do not acknowledge or even respect the array of practices that cultural, heritage and scientific institutions are already undertaking. Attempts to solve accounting "problems" in this arena that do not involve such institutions in the development of their own sector's accountability framework are likely to remain problematic. At the time of writing, the IPSASB has not published any specific guidance or advice on the outcomes pertaining to the CP submissions, that were to be lodged by 20 September 2017. This is not necessarily seen as a positive sign in setting out to work on building coalitions in addressing this seemingly perennial accounting controversy.
The paradox of accounting for cultural heritage The study provides strong, time-extended evidence of how the responses to a paradox can evolve over time. The recent framework developed by CAMD and the related shift towards "mutual adjustment" can be understood only if considered as part of a chain that included, in chronological order, "splitting" and partial "opposition", further "splitting" and "assimilation", "splitting" failure and "suppression". These responses also influenced the dynamic nature of the paradox, from latent to salient and back. Prior processual research on paradox management has been conducted mainly at a micro-level through single case studies (Amyar et al., 2019) and has underlined the role of organisational routine (Jarzabkowski et al., 2013) as well as dynamic capabilities (Smith and Lewis, 2011) in influencing whether responses to a paradox at an organisational level are active or defensive.
By means of examining paradox responses at an organisational field level consisting of 16 major cultural institutions, this study complements prior research by addressing the role of coalition building as a contingent element in shaping the responses to a paradox. In fact, this study shows that while responses were unco-ordinated, institutions reacted with a mix of defensive responses such as "splitting", "opposing" and "suppressing". On the other hand, a transition from "suppression" (defensive response, phase 3) to "mutual adjustment" (active response, phase 4) took place when the directors of Australian museums, by means of their umbrella organisational association, approached the issue of financial valuation collectively, thereby increasing their power, or at least their perceived power, to negotiate with accounting standard setters and other regulatory or supervisory agencies.

Conclusions and forward research agenda
What practices and trends in the mandated financial reporting of collections of selected Australian cultural institutions have occurred given the diversity of views, and related controversy, between accounting and curational approaches to valuing the collections in monetary terms? The study shows that apart from initial opposition by a few museums, most of the efforts of Australian cultural organisations have been directed towards keeping the logics of cultural and accounting valuations separated. This distance, two-states of play approach also seems to inform the more recent CAMD-led attempt at introducing a common framework to the monetary heritage valuation of collections by museums in Australia. At the same time, the study illustrates the auditors' preference for an "assimilation" approach, implying tolerance for sudden changes and inconsistencies in the accounting practices. The unwillingness of accounting standard setters, as part of the organised accounting profession, to answer directly to the six questions raised by Wolnizer (1995, 1999) (see also Woon et al., 2019), reflects a similar response, with ongoing ramifications for the perceived accountability of the international accounting profession.
Suggestions for further research can be made at an Australian and global level. Relating to Australia, what lies ahead from phase 4? Will the framework developed by CAMD, or variations in that framework, succeed in standardising the landscape of Australian museums' collections monetary valuation practices, modifying its long-standing "chaotic accounting margin" condition? Will the implementation of the framework lead to synergies between valuation approaches or will unintended consequences lead to a new range of defensive responses? These questions can be addressed by complementing documental analysis with interviews with heritage professionals and accountants working in cultural organisations. Interviews may clarify how the paradoxical tensions between different approaches to valuation unfold both at the individual levelthat is, contradictions in how heritage professionals and accountants each appreciate and understand value and perform their roles accordingly (Lewis, 2000;L€ uscher and Lewis, 2008) and the group levelthat is, contradictions in terms of identity and membership (Schad et al., 2016).
Globally, continuous attention to the study of the risks associated with monetary heritage valuation is required. Additional research is encouraged on the use of language to alter public perceptions of the fundamental values of collections. Based on the research undertaken, the authors believe, on anecdotal grounds, that the term "heritage assets" has become more regularly used in discussing and debating "heritage". An underlying "creeping effect" in the use of language combined with related varying reflections on collections, especially towards an appreciation of collections as financial resources available for potential extraction from public custodianship, is perceived as contrary to traditional cultural heritage values being understood and appreciated in strictly non-financial terms.
Cultural institutions for decades and centuries have placed prime importance on their raison d'etre of developing collections of artefacts held for the purposes of preservation, conservation and education. Altering perceptions across time and space of the non-financial values of cultural heritage, especially under prevailing crisis conditions due to the COVID-19 global pandemic, which has already led some cultural institutions to deaccession part of their collections [11], is a key means of stimulating "cultural value change" (Ellwood and Greenwood, 2016). It is trusted that any move to require cultural institutions to recognise their collections as "assets" in accounting parlance and, accordingly, to place monetary values on their unique collections for financial reporting purposes, does not become the creation of a ready-to-use sales price list for intensive use during the COVID-19 induced economic downturn (Gair, 2020;Hicks, 2020;Kotarba-Morley, 2020). Such "austerity measures" [12] around the globe, as they would assumably be described, are foreshadowed to be accompanied, in due course, with considerable international community regret and further criticism on the accountability of accounting In the authors' perception, this key risk to the very existence of collections is not what accounting standard-setters appear to perceive as important. Therefore, it will be interesting to observe how the IPSASB views this study and many others published earlier and in future of the genre. Will the international accounting profession, represented by IFAC, especially in this domain by the IPSASB, prove to be sufficiently accountable, or not, in making its next move on this now longstanding international accounting controversy? Notes 1. With heritage experts, we refer to professionals in an appropriate branch of conservation, art history and archaeology. We include in this category curators and museum directors. Although the latter may have a managerial training and background, they are responsible for the management of the careful preservation and conservation of the collections. Besides, they are unlikely to be qualified as, or to define themselves, as professional accountants.