Do heritage recognition and disclosure matter? Evidence from university students Open Access

The treatment of heritage items in the General Purpose Financial Reports (GPFRs) of governments has been the subject of an intense scholarly and professional debate ever since Mautz (1981) claimed that the Washington Monument should be qualified not as an asset, but as a liability, citing the negative net cash flows that it generates.

In their seminal book, Jones and Pendlebury (2000, pp. 125–126), while examining public sector accounting from a theoretical point of view, assert that financial reports “above all must be useful” to address users’ needs. In this respect, the literature has generally focused on two main purposes: “rendering accountability to various stakeholders and supporting decision making” (van Helden and Reichard, 2019, p. 480). Similarly, the International Public Sector Accounting Standards (IPSAS) Conceptual Framework states that public sector GPFRs must be “useful to users of GPFRs for accountability purposes and for decision-making purposes” (IPSASB, 2014, paragraph 2.1).

For heritage items, significant contention exists regarding whether they meet the definition of assets and whether their recognition, measurement, and disclosure actually contribute to accountability and decision-making (Aversano et al., 2019, 2020; Aversano and Christiaens, 2023; Biondi and Lapsley, 2014; De Wolf et al., 2020; Noaman et al., 2018a). In fact, even the definition of what qualifies as heritage is controversial.

Significantly, the existing debate has largely been normative. There is a dearth of empirical evidence (Aversano and Christiaens, 2014; Biondi and Lapsley, 2014; Ellwood and Greenwood, 2016; Ferri et al., 2021; Napier and Giovannoni, 2021; West and Carnegie, 2010) on the effects of heritage recognition, measurement, and disclosure for GPFR users. In particular, quantitative evidence is scant as most empirical papers have relied on document analysis or case studies. This paper aims to address this gap by providing preliminary quantitative validation of some existing claims.

Specifically, the paper focuses on university students, investigating whether heritage recognition and extensive disclosure may influence their interpretations of public entities’ financial reports and their subsequent behaviors. Although not representative of the general population, student samples provide a first step toward a better understanding of how the recognition and disclosure of heritage assets may influence citizens’ perceptions (Bouwman and Grimmelikhuijsen, 2016; Hvidman and Andersen, 2016). The IPSAS Conceptual Framework (IPSASB, 2014, paragraph 2.5) states that “citizens are primary users of GPFRs”. Admittedly, whether citizens have an interest in public sector GPFRs is questionable (Jones, 1992). This may also explain why none of the existing empirical studies on the accounting treatment of heritage items has specifically examined citizens or particular segments thereof, such as students. At the same time, under-researched user groups can be a fruitful research avenue (van Helden and Reichard, 2019).

To this end, an artefactual survey experiment was conducted among Italian university students. Italy is the country with the highest number of UNESCO World Heritage sites worldwide. It has also been characterized by an extraordinary historical and cultural continuity. Experiments have gained prominence in the public administration literature (Bouwman and Grimmelikhuijsen, 2016; Li and Van Ryzin, 2017) as they allow to isolate cause-and-effect relationships. The manipulated variables were the recognition and the extensive disclosure of heritage items in the GPFR of a municipality. Students’ interpretations of financial position and performance were investigated by examining their perceptions of the economic, cultural and social value of heritage, their attitudes towards the sale of heritage items, their perceptions of the municipality’s solvency, their acceptance of conservation and promotion costs, and their propensity towards donations and coproduction.

The study contributes to the literature in several ways. First, it contributes to the limited empirical evidence on the accounting treatment of heritage items by testing some of the claims put forward in the academic and professional literature. Second, it attempts to disentangle the separate and combined effects of recognition and extensive disclosure. Third, by focusing on university students, it establishes a foundation for a broader research agenda that seeks to investigate citizens, who are generally presented as primary users of GPFRs, but tend to be largely overlooked by both research and practice. Finally, it does so by adopting a quantitative experimental approach.

The paper is structured as follows. Section 2 reviews the literature. Section 3 develops the hypotheses. Section 4 illustrates data and methods. Section 5 presents the results. Section 6 discusses the results and concludes the paper.

Heritage has proven difficult to define. Rather than attempting a definition, the accounting literature has identified a range of distinctive characteristics (Adam et al., 2011; Barton, 2000, 2005; Mautz, 1988): heritage is characterized by historic, artistic, architectural, aesthetic, scientific, technological, geophysical, or environmental features; governments have a duty to preserve it for the benefit of current and future generations; it is commonly made available to the general public, and its benefits flow directly to the public rather than to the government entity managing it; its preservation and promotion generate costs that user charges cover only to a very limited extent; its use and disposal are restricted; it is irreplaceable; its useful life is long and often indefinite; it usually has no relevant acquisition cost or market price; and its cultural and social value is not fully captured by its economic value (Anessi Pessina et al., 2020; Carman et al., 1999). In its Consultation Paper on “Financial Reporting for Heritage in the Public Sector”, the International Public Sector Accounting Standards Board (IPSASB) (IPSASB, 2017) defined heritage in accordance with the UNESCO (1972). In the new IPSAS 45 on property, plant and equipment (IPSASB, 2023), however, the IPSASB has reverted to simply providing authoritative guidance on the distinguishing characteristics of heritage.

The unique characteristics of heritage raise significant questions about whether heritage items (i.e. individual objects that represent heritage) should be recognized as assets in governments’ GPFRs. These questions have generally been addressed by discussing whether heritage items meet the definition of assets as developed by accounting standard setters.

Critics of recognition initially argued that heritage items do not qualify as assets because they do not produce economic benefits (Christiaens et al., 2012; Mautz, 1981, 1988). To address this fundamental critique, public-sector standard setters modified the definition of assets by introducing a reference to “service potential” (Anessi Pessina et al., 2022; Micallef and Peirson, 1997). In the IPSAS Conceptual Framework, for example, an asset is “a resource presently controlled by the entity as a result of past events”, with a resource being “a right to either service potential or the capability to generate economic benefits, or a right to both” (IPSASB, 2014, paragraphs 5.6–5.6A). Nevertheless, some academics (e.g. Christiaens, 2022, p. 478–479) remain convinced that “capital goods are fixed assets only if their given status implies the generation of economic benefits”.

Control is the other key issue of contention. In the IPSASB Conceptual Framework (IPSASB, 2014, paragraph 5.11), control “entails the ability of the entity to use the resource (or direct other parties on its use) so as to derive the benefit of the service potential or economic benefits embodied in the resource in the achievement of its service delivery or other objectives”. Critics of recognition (Barton, 2000, 2005; Carnegie and Wolnizer, 1995; Christiaens, 2022) argue that the benefits stemming from the service potential of heritage items flow directly to the community rather than to the government entity. They also claim that treating heritage items as assets would misleadingly inflate net worth, since their sale or transfer is typically prohibited by legal or ethical constraints, limiting their usefulness in meeting liabilities (Carnegie and Wolnizer, 1995; Ouda, 2021). Conversely, the IPSASB (IPSASB, 2017, paragraphs 3.3, 3.7) has highlighted the presence of control by noting that a government entity might “(i) decide whether to set an entrance fee to a museum and deny access to those who do not pay the fee; (ii) prohibit the use of a public square for commercial purposes; or (iii) grant other entities limited reproduction rights to a heritage film or audio-recording” and, more generally, direct the use of heritage items to achieve its objectives which may include “(a) providing access to heritage items directly to individuals (for their education, appreciation, etc.); (b) holding heritage items indefinitely in a custodial capacity; (c) preserving heritage items to benefit the whole community; or (d) promoting heritage-related tourism”.

Supporters of recognition (Micallef and Peirson, 1997; Rowles, 1991, 1992) have claimed that the treatment of heritage items as assets is essential for representational faithfulness. According to Hone (1997), recognition enables the assessment of governments’ and individual public managers’ performance. It also facilitates a socially and economically responsible allocation of public resources. Along similar lines, the European Public Sector Accounting Standards (EPSAS) Working Group (EPSAS, 2017, p. 3) notes that “governments’ expenditure on heritage preservation is in competition with other priorities”. Therefore, heritage recognition is expected to support “decisions on preservation and maintenance of heritage”, “help to prevent destruction, theft and neglect”, and meet citizens’ need “to know whether heritage assets are being cared for and whether resources applied are adequate to ensure heritage assets’ security, preservation and maintenance” (p. 17). Similarly, Biondi et al. (2021), in their analysis of the documents relating to the 2017 IPSASB Consultation Paper on heritage, claim that recognition provides significant benefits for accountability and decision-making. In fact, the IPSASB as well as various national standard-setters (such as the UK with FRS-30, the US with SFFAS 29, New Zealand with FRS-3, Spain with IGAE Resolution of 22/02/2016, South Africa with GRAP 103, and Italy with ITAS 4) are increasingly leaning towards heritage recognition (Aversano, 2015). The IPSASB, in particular, has concluded that “(a) recognition of heritage assets will increase the transparency of heritage-related financial information so that users are better able to hold entities accountable for their heritage-related decisions, particularly those that support heritage preservation; (b) their heritage nature does not prevent heritage items being assets for financial reporting purposes” (IPSASB, 2023, paragraph BC.13). For critics of recognition, however, similar purposes could be achieved through appropriate disclosure, either in the notes or outside the government entity’s financial statements (Barton, 2000; Christiaens, 2022; Stanton and Stanton, 1997).

Measurement entails determining the monetary amounts at which the elements of the financial statements are to be recognized and carried. Debates and guidance on measurement generally focus on the selection and implementation of appropriate measurement models, bases, and techniques.

With respect to heritage, however, measurement has largely been discussed as a condition for recognition, so much so that Micallef and Peirson (1997) suggested that issues of measurement were being confused with issues of recognition. To be recognized as assets in GPFRs, items must not only satisfy the relevant definition, but also be measurable in a way that “achieves the qualitative characteristics and takes account of constraints on information in GPFRs” (IPSASB, 2014, paragraph 6.2). For heritage, the literature has extensively highlighted that the existing measurement models and bases do not adequately capture service potential (Anessi Pessina et al., 2020; Aversano and Christiaens, 2014; Barton, 2000, 2005; Biondi and Lapsley, 2014; Biondi and Oulasvirta, 2023; Carnegie and Wolnizer, 1995, 1999; Harris and Murti, 1998; Hooper et al., 2005; Stanton and Stanton, 1997; Ström, 1997). Moreover, assigning a monetary value to heritage has been described as emphasizing the commercial value of these items rather than their social and cultural significance (Glazer and Jaenicke, 1991; Napier and Giovannoni, 2021; Rentschler and Potter, 1996).

Barton (2005) uses the economic theory of public goods to contend that the inclusion of heritage “in a statement of financial position […] is both irrelevant and unreliable”, where “the provision of irrelevant and unreliable information frustrates good management of the assets rather than enhances it” (p. 439). This is because “there are no active markets for most public heritage assets” and “it is thus not normally possible to obtain a ‘fair market’ value”. Even for items where private markets exist, such as popular works of art, the prices paid will not encapsulate social benefits, but only the private benefits expected to be enjoyed by the buyer (p. 436).

Ellwood and Greenwood (2016) draw on Shrödinger’s cat paradox to analyze two English cases, namely the National Portrait Gallery and Tower Hamlets Council. Their findings show that placing an economic value on heritage items may obscure their social and cultural value and that, ultimately, “it changes reality”. They also mention several significant undesirable effects. In particular, governments may become more prone to sell their heritage. In addition, potential donors may be given the misleading impression that governments holding heritage are wealthy institutions, which could reduce incentives to provide donations.

Napier and Giovannoni (2021) draw on Baudrillard (1994) and Latour (2004) to present a longitudinal study of the accounting treatment for the collection of pictures donated by Thomas Holloway to Royal Holloway College. The study shows that, following recognition, the College faced a tension between treating the collection as an endowment investment that could be sold if necessary and acknowledging its economic or cultural benefits.

Carnegie and Wolnizer use the theoretical lens of accountability to characterize the accounting treatment of heritage as an “accounting fiction” (Carnegie and Wolnizer, 1995) and an “accountability mirage” (Carnegie and Wolnizer, 1996). They highlight that “the accountability of museum managers lies outside the market [and] extends across an array of financial and non-financial responsibilities associated with the acquisition, protection, preservation, conservation and presentation of collection items” (Carnegie and Wolnizer, 1996, p. 371). On this basis, they contend that the financial measurement of collection items is not “useful for enhancing and judging the accountability of those who manage not-for-profit arts institutions having non-commercial objectives” nor “for gauging the financial efficiency with which a public (grant-dependent) arts institution is managed” (Carnegie and Wolnizer, 1995, p. 44). Consistent with these claims, West and Carnegie (2010), in their analysis of Australian public universities, depict the financial reporting of library collections as a “chaotic margin” and problematize the reliability and usefulness of such information. More recently, Carnegie et al. (2022) reiterate that the accounting treatment of heritage creates misrepresentation.

Biondi and Lapsley (2014) examine whether accounting can foster or enhance good governance through the lens of transparency. Their analysis of heritage-related exposure drafts, consultations on technical policy guidance, and formal requirements of accounting policy in the UK shows that accounting for heritage is unlikely to achieve even the simplest level of transparency “mainly due to deep seated, pernicious problems of asset recognition and valuation” (p. 146). In particular, “the act of valuing these assets may appear to achieve transparency, but the levels of subjectivity in valuation serve to obfuscate rather than make financial positions clearer” (p. 152). Similarly, Aversano and Christiaens (2014, p. 170), adopting a user needs approach, administered a questionnaire to politicians in Italian local government and concluded that “the necessity to respond to the information needs of these users, clashes with the difficulties of definition, measurement and recognition of heritage assets”.

Taken together, this literature depicts a paradoxical situation where attempts to assign monetary values to heritage, though intended to improve decision making and accountability, may ultimately undermine those very objectives, as also found by Ferri et al. (2021) in their analysis of annual reports and other documentary sources of Australian public universities.

Once again, a contrasting view is put forward by Micallef and Peirson (1997, p. 33), who argue that the provision of whatever historical cost information is available would “be better than no information about those assets”. IPSAS 45 also provides for heritage items to be recognized as assets when they satisfy the recognition criteria. At the same time, the IPSASB has “agreed with constituents that heritage assets may present measurement difficulties which prevent their recognition” (IPSASB, 2023, paragraph BC.30) and, consequently, included “a requirement for additional disclosures on heritage property, plant, and equipment that is not recognized because it cannot be measured reliably” (IPSASB, 2023, paragraph BC.33).

Disclosure is acknowledged as essential by both supporters and critics of heritage recognition as it allows to address the issues of incomplete information and unreliable measurement (Carnegie and Wolnizer, 1999). Pallot (1990, 1992) highlights the need for non-financial measures of performance in relation to “community assets”, including heritage, which could perhaps be more appropriately presented in a separate “stewardship statement”. Stanton and Stanton (1997) suggest that comprehensive non-financial disclosures provide more accurate and less distorted information. Barton (2000, 2005) argues that such disclosures contribute to effective and efficient asset management. Examples include detailed descriptions of an asset’s condition, restoration needs, usage metrics (e.g. numbers of visitors), and stakeholder satisfaction levels, often presented as “required supplemental stewardship information” (Barton, 2000, p. 15). Aversano and Christiaens (2014) find that politicians prefer qualitative information such as costs of custody, physical condition, and preservation policies.

The need for disclosure persists when heritage is recognized in the statement of financial position (Aversano and Christiaens, 2014; Noaman et al., 2018b). Biondi and Lapsley (2014) advocate a mixed approach that combines financial information with wider information on cultural and societal significance. Botica Redmayne et al. (2023) emphasize the importance of qualitative disclosure in their examination of museums’ financial statements in New Zealand. Christiaens et al. (2012) suggest disclosing a range of descriptive details on heritage as well as financial valuations when available and reliable. Woon et al. (2019) propose a reporting model including financial and non-financial performance indicators and hypothesize that it may encourage citizens to engage in coproduction activities. Biancone et al. (2020) highlight the potential role of popular financial reports in the provision of financial and non-financial information. Anessi Pessina et al. (2020) list a wide set of disclosures recommended by respondents to the 2017 IPSASB Consultation Paper on heritage (see also Aversano et al., 2020).

Current IPSAS guidance, however, does not include specific disclosure requirements for recognized heritage assets, consistent with the general position whereby “the heritage nature of an item is not, by itself, a reason for special financial reporting requirements” (IPSASB, 2023, paragraph BC.13). For unrecognized heritage, governments are expected to disclose “(a) the difficulties in obtaining a reliable measurement that prevented recognition; and (b) the significance of the unrecognized heritage property, plant, and equipment in relation to delivery of the entity’s objectives” (IPSASB, 2023, paragraph 77).

As highlighted in the literature review, the expected benefits and drawbacks of heritage recognition have been extensively discussed. The literature also suggests that these benefits and drawbacks may be respectively amplified and mitigated by disclosure. However, these expectations have not yet been tested through quantitative analyses (a partial exception being Aversano and Christiaens, 2014), nor have they been empirically investigated with specific respect to citizens as primary users of GPFRs or any relevant subgroups, such as students. The purpose of this paper is to address these gaps by examining the following research questions:

It is worth noting that these questions focus on recognition and extensive disclosure, but do not address measurement in its usual sense – namely, the selection and application of appropriate measurement models and bases, such as historical cost, fair value, or current operational value. This is because, in the context of heritage, measurement is generally assumed to be embedded within recognition as a requirement, as discussed in Section 2.3.

These research questions are investigated with specific reference to seven of the arguments presented in the literature. For each argument, the three research questions are translated into a set of three corresponding hypotheses. For each set, the first two hypotheses address the main effects of recognition and extensive disclosure, respectively, while the third focuses on their interaction. Each hypothesis is signed according to the prevailing expectations expressed in the literature. When no clear expectations exist, the hypothesis is formulated in the null.

Perceived economic v cultural and social value. The literature argues that heritage recognition may emphasize economic value at the expense of cultural and social value. Disclosure, on the contrary, may shed light on cultural and social significance. Thus, the first two sets of hypotheses are as follows:

Attitude towards the sale of heritage. The literature suggests that heritage recognition may produce a more favorable attitude towards the sale of heritage, while extensive disclosure may have the opposite effect. Hence the following hypotheses:

Propensity to donate. The literature suggests that heritage recognition may discourage donations, while extensive disclosure may encourage them. More formally:

Assessment of the government’s ability to meet its liabilities. The literature indicates that heritage items cannot generally be used to repay liabilities. However, if recognized, they inflate reported assets, which may bias perceptions of the government’s solvency. Extensive disclosure, per se, is unlikely to affect these perceptions. When accompanying recognition, however, extensive disclosure may underscore that heritage is inalienable or that its sale is undesirable, thus tempering the overly optimistic perceptions produced by recognition. Hence the following hypotheses:

Acceptance of conservation and promotion costs. In the literature, heritage recognition has been claimed to support a socially and economically responsible allocation of public funds. Extant studies also emphasize the need for transparency regarding preservation, conservation, promotion, and management policies. In this context, recognition may increase public acceptance of the relevant costs. These costs are presented in the statement of financial performance regardless of heritage recognition. Presenting the value of heritage items in the statement of financial position may provide a benchmark against which to assess these costs and clarify their purpose. In the absence of recognition, similar effects may be produced by extensive disclosure. The relevant hypotheses are as follows:

Propensity to coproduce. Coproduction can be characterized as being “(i) voluntary (unpaid) and not coercive or normatively set; (ii) an active rather than passive involvement of users; (iii) focused on public value creation in terms of outcomes as well as of outputs” (Fugini et al., 2016, p. 7). The literature suggests that extensive disclosure can stimulate engagement in heritage-related coproduction. The expected impact of recognition is more ambiguous because two competing forces are seemingly at work. Similar to disclosure, recognition may highlight the significance of heritage. At the same time, it may inflate perceptions of the entity’s wealth, thus reducing incentives to coproduce. Therefore, the final set of hypotheses is as follows:

To test the hypotheses, an artefactual survey experiment was conducted using a 2x2 between-subject design. Experimental studies have gained prominence in the public administration literature (Bouwman and Grimmelikhuijsen, 2016; Li and Van Ryzin, 2017) as they allow to isolate cause-and-effect relationships by creating statistically equivalent groups that differ systematically only for the experimental treatment. This design also supports the balanced distribution of potential biases across groups, thereby strengthening the internal validity of comparisons. Furthermore, in the field of public sector accounting, experiments have been identified as a promising research method to investigate the consequences of budgets and financial reports (van Helden and Reichard, 2019).

The research design was preliminarily tested on small samples of undergraduate and doctoral students, leading to refinements to both the questionnaire and the vignettes. The experiment was subsequently administered online to students enrolled at an Italian university.

Italy is a suitable research setting given its immense heritage, including the world’s highest number of UNESCO World Heritage sites, as well as its extraordinary historical and cultural continuity. It has also recently adopted a public sector accounting standard (ITAS 4) which provides for the recognition of heritage items as assets in the financial reports of government entities.

The use of students admittedly limits generalizability, as student samples are not representative of the general population. For example, students are typically more educated, which may lead to different interpretations or biases compared to the broader population. Moreover, they typically do not pay taxes, which may affect their attitudes toward public finances. Nevertheless, student proxies are fairly common in experimental studies, although their use requires appropriate caution (e.g. Bouwman and Grimmelikhuijsen, 2016; Hvidman and Andersen, 2016). Prior research has also reported promising evidence regarding their external validity. One advantage of student samples is that they are usually more homogeneous than other populations, facilitating the detection of true effects (Morton and Williams, 2010, p. 351). Moreover, Druckman and Kam (2011, p. 41) argue that “student subjects do not intrinsically pose a problem for a study’s external validity” and may, in fact, strengthen experimental realism when participants are exposed to complex information.

The link to the questionnaire was provided on the university’s web-based course-management system and all students were asked to voluntarily contribute to the research. Respondents were guaranteed anonymity. They were also offered an economic incentive consisting in the opportunity to win one of two €25 Amazon vouchers.

Following agreement to participate, respondents were provided with instructions and exposed to a vignette presenting the condensed statement of financial position of an imaginary municipality, supplemented by disclosure on the municipality’s heritage. Information on equity was deliberately omitted because the concept is not intuitive (Haustein and Lorson, 2023) and, if included, may have biased respondents’ perception of liabilities. For increased realism, data were sourced from an actual Italian municipality.

Participants were randomly assigned to one of the four treatment groups, which reinforced external validity. To enhance internal validity and, thus, the ability to make causal conclusions (Seltman, 2018), the experiment was blinded. After treatment, all participants were administered the same set of questions.

The manipulated variables were (1) heritage recognition and (2) extent of heritage disclosure. The vignettes are available in Appendix 1.

Regarding heritage recognition, the amount of property, plant and equipment (PP&E) shown in the statement of financial position for groups 2 and 4 was m€ 100 higher than for groups 1 and 3 (m€ 540 v m€ 440). The notes for groups 2 and 4 specified that PP&E included heritage assets for m€ 100, while the notes for groups 1 and 3 stated that heritage was not recognized in the statement of financial position.

Regarding disclosure, all groups received information about the accounting treatment of heritage as well as a table listing and briefly describing the municipality’s four heritage items. For groups 3 and 4, disclosure was more extensive in that the descriptions also included the number of visitors and explained the items’ significance in relation to the municipality’s objectives ^[1].

The questionnaire was designed to reflect the hypotheses, with each set of hypotheses corresponding to one or more questions. Respondents were asked to provide their answers on a 7-point Likert scale.

For hypothesis sets H1 and H2, the questionnaire contained nine items, developed on the basis of Scott (2006) and Grüb and Martin (2020) to capture the perceived economic, cultural and social significance of heritage. For hypothesis set H3, respondents were asked whether they would support the decision to sell part of the municipality’s heritage. For hypothesis set H4, they were asked whether they would be in favor of their family making a donation to sustain the municipality’s heritage. For hypothesis set H5, they were asked to rate the municipality’s ability to meet its liabilities. Given the technical nature of this question, respondents were offered an “I don’t know” response option. Hypothesis set H6 was translated into two questions regarding the adequacy of given amounts of expenses on conservation (m€ 3, accounting for 2.3% of the municipality’s total expenses) and promotion (m€ 2, accounting for 1.5% of total expenses). Finally, for hypothesis set H7, four questions were developed to capture four different forms of coproduction corresponding to the phases of the service cycle (Bovaird and Loeffler, 2012; Brudney and England, 1983; Loeffler and Bovaird, 2018; Nabatchi et al., 2017). These are co-commissioning, co-design, co-delivery, and co-assessment. Co-commissioning involves strategically identifying and prioritizing public service needs, outcomes, and target users. In the questionnaire, respondents were asked to rate their willingness to contribute ideas for the cultural program of the Civic Museum. In co-design, the experiences and inputs of users and their communities are integrated into the creation, planning, or structuring of public services. In the questionnaire, respondents were asked to rate their willingness to interact with local artists and community members to create an interactive exhibition. Co-delivery refers to the joint provision of public services by state and lay actors. It involves working together to deliver services or enhance their delivery. In the questionnaire, respondents were asked to rate their willingness to volunteer time to extend the Museum’s opening hours. Co-assessment involves the joint monitoring and evaluation of public services. In the questionnaire, respondents were asked to rate their willingness to attend exhibitions and provide feedback through a customer satisfaction survey.

Because H1, H2, H6, and H7 were investigated through multiple survey questions, factor analysis was used to aggregate the responses into a single variable. For both economic value (H1) and cultural and social value (H2), one item was removed as it failed to meet the threshold for factor loading. Cronbach’s alphas all exceeded 0.90, indicating strong measurement consistency and confirming the adequacy of the included items to reliably represent the theoretical constructs under study. For acceptance of conservation and promotion costs (H6), the response scale was reversed so that greater values would correspond to higher acceptance. Finally, for perceived solvency (H5), “I don’t know” responses were treated as missing values.

One hundred and seventy-three students from various programs at an Italian University were enrolled in the experiment. At the end of the questionnaire, they were also asked to provide some background information including age, gender, family residency, level and field of education, and experience with heritage. Manipulation checks were included to assess whether participants had correctly perceived the manipulation of recognition and extensive disclosure. Nineteen respondents failed at least one manipulation check and were excluded, leaving a final sample of 154.

Background characteristics are displayed in Table 1, for the entire sample and for each treatment group. Multiple comparisons tests confirmed that the four groups did not differ at the 0.05 level across any of these characteristics, except Visits to cultural establishments in the previous twelve months and Propensity to spend free time in cultural events. Average age was 25. Males accounted for 43% of the sample. The prevailing fields of education were Economics, finance and management (52%) and Humanities (38%). Graduate and doctoral students accounted for 69% of the sample, with the remaining 31% being undergraduates. The large majority (70%) originated from Northern Italy. Most participants had visited cultural establishments 10+ times (31%) or 4–6 times (24%) in the previous twelve months and their average propensity to spend their free time attending cultural events was 5.3 on a 1–7 scale.

Descriptive statistics for the dependent variables are presented in Table 2, for the entire sample and separately for each group and each condition.

The hypotheses were tested using regression analysis, with Group 1 (no recognition nor extensive disclosure) as the baseline. Heteroscedasticity was detected on the basis of the Breusch-Pagan test and controlled for using robust standard errors.

Each set of hypotheses was tested using four regression models. Model 1 included only the main effects for heritage recognition and extensive disclosure. Model 2 added the interaction term. Models 3 and 4 introduced two subsets of controls, designed to avoid the simultaneous presence of the two variables reflecting experience with heritage, which were highly correlated. Interactions were further investigated using marginal analyses. Table 3 shows the results of Model 2. The other models are presented in Appendix 2.

In interpreting the results, the main effects of recognition and extensive disclosure address RQ1 and RQ2, respectively, while the interaction term pertains to RQ3.

Perceived economic value (H1) is significantly increased by both recognition (p < 0.001) and extensive disclosure (p = 0.004), although the latter’s effect is smaller in magnitude (1.333 v 2.217, p < 0.01). The interaction term is statistically insignificant. Therefore, H1.1 and H1.3 are supported, while H1.2, which was formulated in the null, is rejected.

Perceived cultural and social value (H2) is significantly increased by extensive disclosure (1.946, p < 0.001), while both recognition and the interaction term are statistically insignificant. Consequently, H2.2 and H2.3 are supported while H2.1, which posited a negative impact of recognition, is not.

Attitudes towards the sale of heritage (H3) become significantly more favorable in the presence of recognition (2.642, p < 0.001), but this effect is significantly mitigated when recognition is accompanied by extensive disclosure (−1.577, p = 0.010), as also confirmed by marginal analysis (Appendix 3). Extensive disclosure per se, on the contrary, does not exert the expected decreasing effect. In other words, H3.1 is supported, while H3.2 is not and neither is H3.3, which was formulated in the null.

Propensity to donate (H4) is significantly discouraged by recognition (−1.514, p = 0.005) and encouraged by extensive disclosure (0.911, p = 0.024), while the interaction term is statistically insignificant. Therefore, H4.1, H4.2, and H4.3 are all supported.

Perceived solvency (H5) is significantly inflated by recognition (2.204, p < 0.001) and this effect is not mitigated by extensive disclosure as the interaction term is insignificant. As expected, extensive disclosure per se is also statistically insignificant. In other words, H5.1 and H5.2 are supported, while H5.3, which posited a mitigation, is not.

Acceptance of conservation and promotion costs (H6) is significantly increased by both recognition (p < 0.001) and, to a lesser extent (1.414 v 1.976, p < 0.05), extensive disclosure (p = 0.001), while the interaction term is statistically insignificant. Therefore, H6.1, H6.2, and H6.3 are all supported.

Finally, propensity to coproduce (H7) is significantly increased by both recognition (p = 0.001) and extensive disclosure (p < 0.001), with a larger coefficient for the latter (1.851 v 1.344, p < 0.1), while the interaction term is statistically insignificant. These findings provide support for H7.2 and H7.3, but not for H7.1, which was formulated in the null.

Whenever both main effects are significant, while the interaction term is not (H1, H4, H6, H7), marginal analyses (Appendix 3) confirm the additive, but not synergistic, effects of recognition and extensive disclosure.

As mentioned, the robustness of the results was verified by including two subsets of control variables (Models 3 and 4 in Appendix 2). In addition, the independent variables (recognition and extensive disclosure) were re-coded as −0.5, 0.5 (as opposed to 0, 1) in accordance with Van Ryzin (2020). In all these analyses, the results remained virtually unchanged, with the only exception of the attitude towards the sale of heritage (H3), for which the coefficient for extensive disclosure becomes significantly negative (p = 0.035) when the independent variables are re-coded.

The accounting treatment of heritage has been the subject of an intense scholarly and professional debate. This debate has largely been normative, focusing on the definition of heritage and whether heritage items should be recognized as assets in governments’ GPFRs. Limited empirical evidence has also been published on the effects of heritage recognition and disclosure. The purpose of this study has been to empirically examine selected benefits and drawbacks outlined in prior research by adopting an experimental approach. The experiment engaged university students, with the intent to initiate a broader reflection on whether the accounting treatment of heritage is consistent with GPFR’s dual purpose of rendering accountability to various stakeholders – especially citizens as “primary users of GPFRs” – and informing their decision-making processes.

In the empirical analysis, the expected benefits of recognition were all confirmed. Specifically, recognition was found to exert a significantly positive impact on respondents’ perception of the economic value of heritage (H1) and on their acceptance of the conservation and promotion costs required by heritage (H6). These results support some of the arguments proposed by Rowles (1991, 1992), Micallef and Peirson (1997), Biondi et al. (2021), as well as some national and international standard setters (e.g. EPSAS, 2017). Interestingly, the empirical analysis also shows an increase in the propensity to coproduce (H7), thus contributing to the literature on the role of accounting in modern democracies (Woon et al., 2019).

Importantly, however, the undesirable effects of recognition were also confirmed, with the only exception that recognition was not found to impinge on perceived cultural and social value (H2). Recognition was found to generate a more favorable attitude towards the sale of heritage (H3), reduce respondents’ propensity to provide donations (H4), and inflate their perception of the government’s solvency (H5). These results corroborate the claims put forward by critics of recognition (Biondi and Lapsley, 2014; Ellwood and Greenwood, 2016; Ferri et al., 2021; Napier and Giovannoni, 2021).

While the results regarding recognition are problematic, those concerning extensive disclosure are more promising. Similar to recognition, extensive disclosure was found to positively affect respondents’ perceptions of the economic value of heritage (H1), their acceptance of the conservation and promotion costs required by heritage (H6), and their propensity to coproduce (H7). Unlike recognition, extensive disclosure was found to also increase the perceived cultural and social value of heritage (H2) as well as the propensity to donate (H4). Regarding propensity to sell (H3), extensive disclosure per se did not have a significant effect; in the presence of recognition, however, extensive disclosure was found to partially offset the associated increase in propensity to sell. Only the overestimation of solvency (H5) resulting from recognition did not appear to be mitigated by the provision of extensive disclosure. This evidence aligns with prior literature as both supporters and critics of heritage recognition acknowledge the importance of providing qualitative information (Anessi Pessina et al., 2020; Aversano and Christiaens, 2014; Barton, 2000; Biondi and Lapsley, 2014; Botica Redmayne et al., 2023; Stanton and Stanton, 1997; Noaman et al., 2018b; Woon et al., 2019). Importantly, the results also distinguish the standalone effects of extensive disclosure from those generated through its interaction with recognition.

Overall, the results of this study thus suggest that recognition and extensive disclosure may not contribute equally to heritage-related accountability and decision-making. Recognition appears to produce a mixed pattern of effects, while extensive disclosure is associated with a more consistently positive impact.

With respect to accountability, the results indicate that recognition can highlight the economic relevance of heritage and increase acceptance of the costs associated with its conservation and promotion. In this way, recognition may enhance the visibility of heritage and of the preservation and valorization policies implemented by the relevant government entity. At the same time, however, the results reveal that recognition may undermine accountability in important respects. Most significantly, recognition appears to inflate perceptions of government solvency. This is a particularly critical issue, as accountability relies on stakeholders developing a realistic understanding of the government’s financial position and its capacity to meet its obligations. In addition, the more favorable attitude toward the sale of heritage induced by recognition raises concerns about the possible commodification of items that should be held primarily for intergenerational, cultural, and social purposes.

Similar tensions emerge with regard to decision-making. On the one hand, recognition may encourage a more active relationship between citizens and public institutions through co-production. It also appears to increase support for spending on heritage conservation and promotion. On the other hand, it appears to discourage donations and increase support for sale. These latter effects are problematic because they may distort decisions and behaviors that are important for the long-term preservation of heritage.

The evidence on extensive disclosure is more encouraging. The results suggest that it better captures and conveys the multidimensional nature of heritage and is, therefore, more closely aligned with both dimensions of GPFRs’ dual objective. The only important exception relates to the overestimation of government solvency, which disclosure does not appear to correct.

Overall, these results contribute to the scholarly and professional debate in several ways. First, they provide empirical evidence regarding the benefits and drawbacks of heritage recognition and extensive disclosure. Accounting “must above all be useful” (Jones and Pendlebury, 2000, pp. 125–126). Therefore, the recognition of heritage items as assets in GPFRs is fully justified only if it contributes to accountability and decision-making. Significantly, the findings of this study suggest that extensive disclosure may bring benefits similar to recognition without generating the same drawbacks. Second, the results shed some light on the separate and combined effects of recognition and extensive disclosure. Some benefits of extensive disclosure cumulate with those of recognition, while others are lost in its presence. Extensive disclosure also produces benefits that recognition does not provide and moderates some drawbacks of recognition. Third, the study establishes a foundation for a broader research agenda that seeks to investigate the perceptions and intended behaviors of citizens. Citizens are generally presented as primary users of GPFRs, but they tend to be largely overlooked by both research and practice. In the case of heritage, investigating citizens is particularly important because heritage is predominantly held for the benefit of present and future generations. The literature has extensively questioned whether citizens have an interest in public sector GPFRs (Jones, 1992). Importantly, the findings of this study indicate that, at least among university students, accounting information may significantly affect judgments and attitudes once it is brought to the their attention.

Finally, from a methodological perspective, the study is valuable in that, unlike extant literature, it adopts a quantitative experimental approach, which allows to test causal relationships.

From a practical viewpoint, the results of this study may have interesting implications for both standard setters and GPFR preparers. For standard setters, the most important finding is seemingly the potential role of extensive disclosure. If heritage is not recognized, extensive disclosure may provide important benefits. If heritage is recognized, extensive disclosure may offer additional benefits and also partially mitigate some drawbacks of recognition. Based on these results, guidance on disclosure should be significantly expanded and extensively exemplified. Similarly, preparers should consider that heritage recognition may enhance perceived economic value, acceptance of conservation and promotion costs, and propensity to coproduce, but it may also generate some unintended effects. Therefore, extensive disclosure should not merely accompany recognition, but be strategically designed to emphasize the cultural and social significance of heritage. More generally, preparers should be aware that some groups of citizens may not be completely indifferent to, nor unaffected by, financial reports, once the relevant information is effectively communicated, for example through popular reports (Biancone et al., 2020).

As with any research, this study is not free of limitations, which also suggest caution in practical interpretation. As already acknowledged, a primary concern relates to external validity and generalizability, as the participants were all students from a single Italian university. Despite their widespread use in experimental studies, student samples cannot be considered representative of the general population. A related limitation is the possibility of winning a €25 Amazon voucher, which may have influenced respondents’ answers. However, this was arguably too weak an incentive to jeopardize the validity of the study (Mohr and Kearney, 2021, p. 24). Another related limitation is that Italy’s profound appreciation for its heritage may have affected some responses. More importantly, the need to maintain a manageable vignette required that disclosure be limited to heritage and positioned immediately below the condensed statement of financial position. This is an evident simplification compared to real-life settings, where disclosures about heritage would be part of a lengthy section of notes and, therefore, not immediately visible.

The results and limitations of this study suggest several future research avenues. Second-stage studies are needed that rely on more representative samples, also from other national contexts. The benefits of disclosure should be verified using methodologies that better approximate the complexity of actual financial reports. The possible contents of disclosure could be analyzed in more detail, for example based on the suggestions offered by the literature and the guidance issued by international and national standard setters. The choice of measurement bases, which this study left implicit, could also be explored. Finally, an interesting research avenue could be the relevance of financial information on heritage for internal users.

The supplementary material for this article can be found online.