An empirical comparison of the effect of XBRL on audit fees in the US ...

Journal of Contemporary Accounting & Economics 11 (2015) 89–103

Contents lists available at ScienceDirect

Journal of Contemporary Accounting & Economics j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j c a e

An empirical comparison of the effect of XBRL on audit fees in the US and Japan Yuan George Shan *, Indrit Troshani, Grant Richardson School of Accounting and Finance, University of Adelaide, 10 Pulteney Street, Adelaide, SA 5005, Australia

A R T I C L E

I N F O

Article history: Received 25 March 2014 Received in revised form 12 December 2014 Accepted 20 January 2015 Available online 24 February 2015 JEL: M41 M42 O33

A B S T R A C T

This study examines the effect of eXtensible Business Reporting Language (XBRL) on audit fees, and assesses whether and how this effect varies across different countries. We specifically consider publicly listed firms in the US and Japan over the eight-year period centered around the dates on which XBRL was first mandated for use by publicly listed firms in these jurisdictions. Our comparative results show that XBRL use is inversely associated with audit fees, which are in turn positively associated with firm size. We also find that, overall, XBRL moderates the association between firm size and audit fees in both the US and Japan. However, these moderation effects are weaker among Japanese firms. © 2015 Elsevier Ltd. All rights reserved.

Keywords: XBRL Audit fees Firm size Agency theory

1. Introduction There have been growing calls from investors and regulators around the world to increase the degree of transparency in financial reporting to improve confidence in the capital markets, especially in response to several high-profile corporate scandals at the beginning of the 21st century, such as Enron and WorldCom in the US and Kanebo and Nanaboshi in Japan (Numata and Takeda, 2010; Roohani et al., 2009). Several countries have thus carried out legal reforms and promulgated new directives and laws. For example, the 2002 Sarbanes-Oxley Act (SOX) in the US and the 2006 Financial Institution and Exchange Laws (J-SOX) in Japan recognize the importance of transparency in financial reporting (Bedard et al., 2009; Roohani et al., 2009; TheCorporateCouncil, 2011; Uehara et al., 2008). To a large extent, financial reporting transparency is adversely affected by the lack of interchangeability in the ways in which financial data are collected, processed, repurposed and reported. In fact, the paper and digital formats that are currently used, including HTML, PDF, and Excel, do not provide sufficient semantics to enable the automated processing of financial data, making the realization of efficient and effective transparency objectives expensive and difficult to achieve (Debreceny et al., 2010; Doolin and Troshani, 2004, 2007; Troshani and Doolin, 2007; Troshani and Rao, 2007; Troshani et al., 2014). eXtensible Business Reporting Language (XBRL) is a data formatting standard that enables the electronic communication of financial reports (Blankenspoor et al., 2014; Efendi et al., 2014; Troshani and Lymer, 2010). It can enhance the interchangeability of financial data by streamlining and integrating information flows among heterogeneous organizations. XBRL can facilitate the exchange of financial data between different computer platforms and accounting information systems

* Corresponding author. School of Accounting and Finance, University of Adelaide, 10 Pulteney Street, Adelaide, SA 5005, Australia. Tel.: +61 8 8313 6456; fax: +61 8 8313 0170. E-mail address: [email protected] (Y.G. Shan). http://dx.doi.org/10.1016/j.jcae.2015.01.001 1815-5669/© 2015 Elsevier Ltd. All rights reserved.

90

Y.G. Shan et al./Journal of Contemporary Accounting & Economics 11 (2015) 89–103

and in doing so, enhance the transparency of financial reports (Abdolmohammadi et al., 2002; Doolin and Troshani, 2007; Troshani and Lymer, 2010, 2011). With greater transparency, it is reasonable to expect that firm monitoring functions, including external auditing, could be enhanced and related auditing costs reduced. The objective of this study is to examine the effect of XBRL on audit fees, and to assess whether and how this effect varies across different jurisdictions. We specifically focus our attention on external auditing because it is considered to be an important monitoring function in the business information supply chain (Simkin et al., 2012). In fact, auditing evaluates the accuracy and fairness of financial statements relative to a jurisdiction’s generally accepted accounting principles (GAAP) (De George et al., 2013; Simkin et al., 2012) and consequently improves the credibility of the firms which produce them. This makes it less costly for firms to achieve their investment growth aims (Khurana and Raman, 2004) while also facilitating compliance with appropriate legislation (Li et al., 2007; Rezaee, 2009). We argue that XBRL enhances the transparency of financial reporting, which facilitates external auditing functions. With greater transparency, audit risks are likely to be reduced, resulting in lower audit costs for the firm (Khadaroo, 2005). Although XBRL is expected to “revolutionize” business reporting (Abdolmohammadi et al., 2002), there has been little research on its related effects. Specifically, research has examined the effects of XBRL on alleviating information asymmetry (Yoon et al., 2011), improving the information environment by reducing event return volatility and cumulative abnormal returns (Bai et al., 2013) and information content (Efendi et al., 2009, 2014), reducing the cost of equity capital (Li et al., 2012), improving corporate governance (Premuroso and Bhattacharya, 2008) and enhancing the forecast accuracy of financial statement analysts (Liu et al., 2014). To the best of our knowledge, no study in the accounting literature has specifically examined the effect of XBRL on financial statement auditing. We conduct pre- and post-XBRL adoption comparisons of the audit fees paid by publicly listed firms in two different jurisdictions: the US and Japan. We deliberately select these particular jurisdictions because the use of XBRL for filing financial statements has been mandated for all publicly listed firms in the US and Japan with the expectation of greater transparency and timely access to financial statement information, resulting in more efficient and effective regulatory processes, including auditing (Starr, 2012). In December 2008, the US Securities and Exchange Commission (SEC) mandated that publicly listed firms file key financial statements in XBRL format using a phased-in approach, commencing in June 2009 (Srivastava and Kogan, 2010). In its efforts to achieve international interoperability with the International Accounting Standards Committee Foundation (IASCF) and the SEC, Japan’s Financial Services Authority (FSA) mandated that all publicly listed firms file their financial statements in XBRL format, commencing in April 2008 (Bai et al., 2013; FSA, 2008; Kobayashi, 2008). Given that firms in these countries are subject to regulatory mandates for XBRL use, the US and Japan represent ideal settings for conducting our research. Moreover, both countries constitute important global jurisdictions characterized by a large number of statements filed by firms of various sizes, suggesting that these jurisdictions might offer greater insight than other less prominent jurisdictions where XBRL has also been mandated (Bai et al., 2013). We use a data set containing 17,010 firm-year observations of firms listed on the New York Stock Exchange (NYSE) and NASDAQ in the US over the 2005–2012 period, and a data set comprising 7067 firm-year observations for firms listed on the Tokyo Stock Exchange (TSE) in Japan over the 2004–2011 period. Thus, we are able to capture four years of observations before and after the XBRL use mandates took effect in the two jurisdictions. Finally, we focus our attention on publicly listed firms because they are subject to the XBRL use mandates in the US and Japan. We specifically compare and contrast the US and Japan, given the paucity of research concerning whether the effect of XBRL on auditing in a common law jurisdictional setting (e.g. the US) can be generalized to non-common law jurisdictional settings (e.g. Japan), where country differences exist and legal regimes may offer less investor protection and weaker laws for suing auditors for misconduct and negligence (Francis, 2004; Markus and Kitayama, 1991). Our comparative results indicate that XBRL filers in the US and Japan have experienced savings in audit costs as measured by audit fees paid to external auditors. In particular, XBRL use is negatively associated with audit fees, which are, in turn, positively associated with firm size. We also observe that, overall, XBRL moderates the association between firm size and audit fees both in the US and Japan. However, these moderation effects are weaker among Japanese firms. Our study’s contribution is threefold. First, to the best of our knowledge, this is the first cross-country comparative study to consider the effect of XBRL on audit fees. In fact, we use large data sets with eight-year time spans for the US and Japan, thereby enhancing the generalizability of our findings. Furthermore, we also observe differences regarding the effect of XBRL between the US and Japan, which are characterized by very different legal frameworks (La Porta et al., 2003). Second, XBRL is a complex technology that has raised a number of concerns in relation to the compliance costs and benefits of regulatory business reporting (Li et al., 2012). Although XBRL reporting is expected to enhance transparency and facilitate auditing, it suffers from significant setup, implementation, education and training costs (Bai et al., 2013). Our study contributes to the ongoing debate over whether the purported economic benefits of XBRL do indeed materialize in practice (Efendi et al., 2014; Hwang et al., 2008; Troshani and Lymer, 2010). We provide evidence which suggests that XBRL assists firms to become more transparent to the public and therefore easier and less risky to audit, resulting in lower auditing fees. This study thus contributes to the broader literature concerning the economic implications for auditing that result from enhanced financial reporting practices, while also responding to calls to examine the economic viability of XBRL (Debreceny et al., 2010). Third, this study has implications for auditors and accountants in countries where XBRL adoption has been mandated and in those countries where XBRL is currently being considered for adoption. The existing literature suggests that policy-making deliberations and regulations in the accounting realm are predominantly driven by political rather than scientific objectives (see e.g. Francis, 2004). Our findings can thus provide a solid foundation to facilitate a change in this tradition, at least as it pertains to XBRL use. More specifically, our findings can inform the business case for using XBRL in the regulatory sphere.


91

The remainder of this paper is organized as follows. Section 2 discusses the theoretical framework and develops our hypotheses. Section 3 describes the research design and data collection process. Section 4 reports and discusses our empirical results. Finally, Section 5 concludes the paper. 2. Theory and hypotheses development This study is underpinned by agency theory, according to which a firm consists of a nexus of contracts between its principals (i.e. owners of economic resources, including shareholders) and its agents (i.e. managers who are responsible for using and controlling these resources) (Jensen and Meckling, 1976). Agency theory posits that principals do not have access to all available information when decisions are made by agents. This information asymmetry can generate an “adverse selection” problem that impairs the principal’s ability to determine whether agents are, in fact, acting in the best interests of the firm (Jensen and Meckling, 1976). Additionally, agents may face the “moral hazard” of acting against the principal’s interests (Scapens, 1985) while attempting to maximize their own wealth (Sarens and Abdolmohammadi, 2011), which can be manifested in the form of excessive use of perquisites, asset misappropriation and enhancement of salary (Christie and Zimmermann, 1994; Rediker and Seth, 1995). Taken together, these problems give rise to agency costs, which the principal attempts to mitigate by establishing monitoring processes such as external auditing (Francis and Wilson, 1988). Auditors have the responsibility to verify and ensure that agents act according to the principal’s best interests (Helliar et al., 1996; Nikkinen and Sahlström, 2004; O’Sullivan, 2000). Auditors accomplish this responsibility by monitoring and verifying the actions undertaken by management through financial reporting, internal controls, and risk management (Sarens et al., 2009), thereby fulfilling a “policing” function as part of the regulatory processes in financial reporting (Alleyne et al., 2013). The value of external auditing results from the expectation or probability that an auditor will detect and/or report significant breaches (or risks thereof) in the financial statements produced by the firm’s management (e.g. material omissions or misstatements) and accounting systems (e.g. inadequate internal controls) (DeAngelo, 1981; Dopuch and Simunic, 1982; Francis, 2004). Jensen and Meckling (1976, p. 329) claim that the “existence and size of the agency costs depend on the nature of [agent] monitoring costs.” Leventis et al. (2011, p. 113) also argue that “a specific and measurable agency cost” is captured by audit fees. Specifically, a reduction in audit fees signals that agency costs are reduced (Leventis et al., 2011). Audit fees thus represent an important proxy of agency costs and there is a great deal of evidence to show that audit fees are affected by agency costs (Gul, 1999; Gul and Tsui, 2001; Leventis et al., 2011). For instance, auditors are likely to spend much more time inspecting managers’ activities in cases where agency problems are suspected (Dopuch, 1980; Dopuch and Simunic, 1982; Leventis et al., 2011; Nikkinen and Sahlström, 2004). Proponents of XBRL assert that it can produce greater corporate reporting transparency by reducing the cost of collecting, processing and compiling financial information, and by streamlining firms’ internal and external financial reporting (Premuroso and Bhattacharya, 2008). In turn, this facilitates auditing processes related to the identification of material omissions and misstatements (Khurana and Raman, 2004). Specifically, financial reports in XBRL form can be used for quality surveillance and reporting integrity (Roohani et al., 2009). With XBRL, auditors can easily track the relationship between accounts in the financial reports and internal control evaluation in business processes (Du and Roohani, 2007), flexibly evaluate the relationships between internal control objectives, risks and the corresponding control activities (Roohani et al., 2009), and efficiently and reliably run audit trailing tests across different organizations and disparate accounting information systems (Cohen, 2007; Roohani et al., 2009). XBRL can also facilitate auditing, thus allowing auditors to integrate the appraisal of internal controls over financial reporting with the auditing of financial statements. This can result in significant cost savings, including the reduction of both audit engagement costs and the costs of manually performing substantive internal control tests (Rezaee, 2009). The mechanisms by which XBRL can reduce audit fees operate at the auditor level in three ways. First, XBRL improves auditors’ access to and analysis of financial information by facilitating data gathering, integration and sharing between the firm (i.e. the auditee) and the auditor (Boyle et al., 2014). For example, complete audit data can be offered electronically to auditors at the beginning of the audit process, thus addressing the piecemeal data access inefficiencies that characterize the traditional audit process (Brands, 2013). By standardizing the data format, XBRL creates data access efficiencies that were not previously available to auditors who may have had access to financial data electronically, but had to undertake manual retrieval of the data to overcome the lack of interchangeability issues affecting firm data typically available in different formats (e.g. MS Excel spreadsheets, MS Word documents, PDF, and HTML). Additionally, XBRL facilitates the auditor in carrying out analytical reviews during the audit review process, thus reducing labor, time and costs (La Rosa and Caserio, 2013).1 Second, because XBRL enables financial data to be captured at a much more granular level with “detail tagging” (PWC, 2011) for each disclosure item, it facilitates the work of auditors by supporting computerized auditing processes (Bizarro and Garcia, 2010; Eccles and Krzus, 2010). This enables the automatic validation of calculated numbers or compliance with disclosure checklists (Bizarro and Garcia, 2010). XBRL also improves analysis, enhances audit trails and reduces spreadsheet proliferation (Bizarro and Garcia, 2010). Third, XBRL facilitates the audit confirmation process involving the collection of evidence from third parties that corroborates relevant assertions about a firm’s account balances, transactions and disclosures (Janvrin et al., 2010). To illustrate,

1 Early evidence supporting this point indicates that the time required for auditors to review a bank’s quarterly financial statements, for instance, was significantly reduced from approximately 70 days to just two days after XBRL was introduced (La Rosa and Caserio, 2013; Roth, 2009).

92


when financial data are captured in XBRL format, suitable XBRL-enabled applications can assist the audit process by quickly identifying anomalies that might be indicative of financial reporting fraud. Financial data in XBRL format can be swiftly analyzed to identify patterns of risk indicators or potentially faulty or missing internal controls over financial reporting that might suggest that a firm is engaging in aggressive earning management.2 Therefore, it is likely that XBRL can facilitate the auditing process and reduce audit related costs. Based on the aforementioned claims, in addition to the emerging evidence of the benefits of XBRL (see e.g. Boyle et al., 2014; Brands, 2013; Holzinger, 2013; La Rosa and Caserio, 2013; Roselli, 2013; Tysiac, 2013), the main motivation of our study is to empirically investigate whether XBRL does in fact reduce agency costs, resulting in lower audit fees (Leventis et al., 2011). We deliberately focus our attention on publicly listed firms and auditing fees paid to external auditors, which is particularly important given the potential agency costs that can arise as a direct result of the separation of firm ownership and management (Francis, 2004). Prior research has found a significant correlation between audit fees and firm size, although the results are mixed (see e.g. Casterella et al., 2004; Firth, 1985; Fleischer and Goettsche, 2012; Fung et al., 2012; Simunic, 1980; Wallace, 1984). For instance, several studies have found that audit fees are inversely related to firm size (e.g. Simunic, 1980; Wallace, 1984), which is attributable to a firm’s accounting systems and internal audit functions. Specifically, large firms normally benefit from more sophisticated accounting systems and more efficient internal audit functions than small firms. Thus, large firms can facilitate external auditing while also enabling external auditors to use the work carried out by internal auditors to reduce their audit fees. Large firms also have stronger bargaining power than small firms, which enables them to increase the scale discounts on audit fees (Casterella et al., 2004; Fung et al., 2012). In contrast, other studies have found a positive association between audit fees and firm size (e.g. Firth, 1985; Fleischer and Goettsche, 2012), which is attributable to the increased complexity and risk in large firms (Firth, 1985) and the political visibility of large firms relative to their small firm counterparts (Premuroso and Bhattacharya, 2008). The increasing complexity of accounts in large firms can make it more difficult for external auditors to audit and detect fraud (Firth, 1985). External auditors may also charge large firms a premium when they perceive the audit to be risky (Khadaroo, 2005; Reynolds and Francis, 2001). Risk can be driven by potential litigation costs and the loss of goodwill (Firth, 1985). Large firms are also more likely to receive additional public attention because they are more politically visible than small firms (Marston and Polei, 2005; Meek et al., 1995; Premuroso and Bhattacharya, 2008). Thus, large firms often respond to public scrutiny pressures by increasing the extent of information disclosure, which in turn increases the external auditor’s work effort and results in higher audit fees. Based on these arguments, we hypothesize that: H1a. Ceteris paribus, audit fees are reduced because of XBRL. H1b. Ceteris paribus, XBRL moderates the association between firm size and audit fees. Prior research has also examined the effect of a country’s legal regime on auditor behavior through the incentives that emerge as a result of standards that auditors must fulfill to comply with their statutory audit requirements (Francis, 2004). Specifically, forms of punishment for auditor negligence or misconduct that are institutionalized as legal liabilities can incentivize auditors to fulfill their auditing responsibilities to different extents (La Porta et al., 2003). For instance, auditors that operate in countries where the threat of litigation is high (e.g. the US) are relatively more conservative towards their clients because such countries are likely to have legal systems in place that protect investors, such as the ability to sue auditors (Francis and Wang, 2004; Frantz, 1999; Seetharaman et al., 2002). Based on the “insurance hypothesis,” which has received empirical support (see e.g. Brown et al., 2013; Numata and Takeda, 2010), the auditor’s litigation risk is an important construct that can explain the link between audit quality and its economic value. Accordingly, the value of audit quality can be derived by the insurance that is provided by the auditors to cover the potential loss that may result when financial statements are misrepresented; that is, they contain material misstatements and omissions (Numata and Takeda, 2010; Taylor and Simon, 1999). Given that auditors have a legal responsibility to discover and report breaches in their clients’ financial statements and accounting systems (DeAngelo, 1981), investors can recover at least part of their losses due to misrepresented statements by pursuing litigation against their auditors, the value of which determines the audit quality (Beatty, 1993; Khurana and Raman, 2004). Hence, there is a connection between litigation risk and audit fees in that the audit fee will increase as the losses from the imposition of a legal liability increase (Beatty, 1993). Thus, it is possible that XBRL may also reduce audit fees by operating from the investor’s viewpoint concerning the auditor litigation risk. Specifically, as XBRL is expected to offer investors an effective way to prepare, publish, exchange and analyze financial information (Boritz and No, 2008; Hodge and Pronk, 2006; Hodge et al., 2004; Lymer and Debreceny, 2003), it has the potential to improve information disclosure quality and decrease an investor’s information processing costs (Blankenspoor et al., 2014; Efendi et al., 2014; Li et al., 2012;

2 Some examples include the use of specific wording in, or omissions from, management discussions and analysis, high proportions of off-balancesheet transactions, compensation incentives (e.g. bonus stock options) that do not meet analysts’ earnings targets and the understatement or overstatement of earnings based on the level of discretionary (abnormal) accruals (e.g. the use of accounting policies that increase reported book earnings while at the same time minimizing taxable income) (Boyle et al., 2014; Holzinger, 2013; Roselli, 2013; Tysiac, 2013).


93

Yoon et al., 2011). With the greater transparency offered to investors, XBRL can thus empower investors with a greater ability to confirm the accuracy of an auditor’s opinions, which can, in turn, reduce the risk of litigation against auditors. As the auditor’s litigation risk declines, audit fees are also likely to reduce.3 Although US research supports the connection between auditor litigation risk and audit fees (see e.g. Abbott et al., 2006; Lyon and Maher, 2005; Pratt and Stice, 1994; Seetharaman et al., 2002; Simunic, 1980; Simunic and Stein, 1996; Venkataraman et al., 2008), evidence from outside of the US provides mixed results (see e.g. Francis, 1984; Francis and Stokes, 1986; Gul and Tsui, 1998). The auditing contexts in the US and Japan differ in relation to litigation risk. In contrast to the US where litigation against auditors is a major concern (Pratt and Stice, 1994; Seetharaman et al., 2002; Taylor and Simon, 1999), Japan is considered to be a low-litigation country (Ginsburg and Hoetker, 2006; Skinner and Srinivasan, 2012). According to Japan’s social customs, it is preferable that disputes be settled in ways that do not involve legal action through the Courts. This claim is supported by recent evidence showing that investors have rarely sued audit firms for negligence in Japan (see e.g. Numata and Takeda, 2010). However, while there is some evidence that litigation in Japan is increasing, there is also evidence suggesting that litigation costs in Japan remain lower than such costs in the US (Ginsburg and Hoetker, 2006; Skinner and Srinivasan, 2012). Therefore, it is reasonable to expect that Japanese auditors may charge relatively lower audit fees than US auditors because they have relatively lower exposure to litigation (Fukukawa et al., 2006). Based on these arguments, we hypothesize that: H2. Ceteris paribus, the effect of XBRL use on audit fees is higher in the US than in Japan. 3. Research design 3.1. Data The data for this study were collected from the Thomson ONE Banker database. For the US, we include all firms listed on the NYSE and NASDAQ for the eight-year period from 2005 to 2012; that is, four years before and four years after the XBRL use mandate transpired in the US (see Fig. 1). The SEC implemented a “phase-in” approach (Srivastava and Kogan, 2010) and mandated that publicly listed US firms file their financial statements in XBRL format in accordance with their market capitalization (i.e. above USD5 billion from June 15, 2009, between USD700 million and USD5 billion from June 15, 2010, and the remaining firms from June 15, 2011) (SEC, 2009). Thus, for the US firms, the 2005–2008 period is denoted as pre-XBRL and the 2009–2012 period as post-XBRL. After removing voluntary XBRL users4 before 2009 and firms with missing data, our US data set contains 17,010 firm-year observations. The Japanese data set includes all firms listed on the TSE for the 2004–2011 period. Because the Japanese FSA mandated that all publicly listed firms must file their financial statements in XBRL format from April 2008 (Bai et al., 2013; FSA, 2008; Kobayashi, 2008), we consider the four-year pre-XBRL and post-XBRL periods as 2004–2007 and 2008–2011, respectively (see Fig. 1). After eliminating missing data for the variables, the Japanese data set consists of 7067 firm-year observations. 3.2. Model development The regression model used to examine the effect of XBRL on audit fees (H1a) and the moderation effect of XBRL and firm size on audit fees in the US and Japan (H1b) is represented as follows:

AUDITFEESi = α + β1XBRL i + β 2FIRMSIZEi + β 3XBRL i × FIRMSIZEi + β 4DERATIOi + β5ΔEPSi + β6 ΔSALESi + β7 TOBINSQ i 8

11

j =1

k =1

+ β8BIG4i + β9FORASSETi + β10FORSUBi + β11RTARATIOi + β12ITARATIOi + γ j ∑ YEAR i + ηk ∑ INDUSTRYi + ε i Model (1)

To investigate the difference in the interaction of XBRL use and firm size between the US and Japan (H2), we use an expanded model (see Model 2) to estimate the significance of the differences (t-test) in the pairs of estimated coefficients reported in the US and Japanese data sets – β3 (USA) versus β3 (Japan), as follows:

t -difference (USA-JAPAN)

⎧ fUSA (β1XBRL , β 2FIRMSIZE, β 3XBRL ∗ FIRMSIZE, β 4DERATIO, β5ΔEPS, β6 ΔSALES, ⎪⎪β7 TOBINSQ , β8BIG4, β9FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO) i ,t =⎨ f JAPAN (β1XBRL , β 2FIRMSIZE, β 3 XBRL ∗ FIRMSIZE, β 4DERATIO, β5 ΔEPS, β 6 ΔSALES, ⎪ ⎩⎪β7 TOBINSQ , β8BIG4, β9FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO)i,t

Model (2)

3 Early evidence supports the plausibility of this argument. For example, before XBRL was implemented in the US, accounting fraud and improper disclosures constituted more than 25 percent of the SEC’s civil enforcement actions. This figure dropped to 11 percent in 2012, after XBRL use was first mandated (Eaglesham, 2013). 4 Before the XBRL mandate took effect in the US, the SEC accepted voluntary XBRL filing. A number of corporate filers voluntarily filed statements to the SEC in XBRL format (see Premuroso and Bhattacharya, 2008). These filers were removed from the US data set.

94


USA

Pre-XBRL mandate 2004

2006 2005

Japan

Post-XBRL mandate 2008

2007

Pre-XBRL mandate

2010 2009

2012 2011

Post-XBRL mandate

Fig. 1. XBRL adoption mandate timelines in USA and Japan.

3.3. Dependent variable Our dependent variable is represented by audit fees (AUDITFEES), which have been widely used in prior research as a proxy of firm audit costs (see e.g. Boo and Sharma, 2008; Elder et al., 2009; Fleischer and Goettsche, 2012; Leventis et al., 2011; O’Sullivan, 2000; Zaman et al., 2011). In accordance with prior research, we measure AUDITFEES as the natural logarithm of total audit fees at the end of the fiscal year. We note that audit fees are consistently disclosed by US and Japanese publicly listed firms in their annual financial statements. 3.4. Independent variables XBRL and firm size (FIRMSIZE) denote the independent variables of interest in this study. Following Premuroso and Bhattacharya (2008), XBRL is measured as a dummy variable, coded as 1 if the firm is an XBRL filer, and 0 otherwise. This coding was carried out consistently with the phased-in approach that was adopted by the SEC in the US. Consistent with prior research in the accounting and audit literature (see e.g. Abbott et al., 2007; Gul et al., 2007; Lai, 2009; Premuroso and Bhattacharya, 2008), we measure FIRMSIZE as the natural logarithm of total assets at the end of the fiscal year. 3.5. Control variables We define our control variables as follows. The debt-to-equity ratio (DERATIO) is measured as the ratio of long-term debt to total equity (e.g. Seetharaman et al., 2002); change of earnings per share (ΔEPS) is computed as the difference between EPS at year t and EPS at year t–1 (e.g. Chen and Zhang, 2010); change of sales ratio (ΔSALES) is calculated as the difference between sales at year t and sales at year t–1, divided by sales at year t–1 (e.g. Houqe et al., 2012); Tobin’s Q (TOBINSQ) represents a market performance indicator and is computed as the ratio of the market value of stock and the book value of debt divided by the book value of total assets (e.g. Shan, 2014); Big 4 auditor (BIG4) is a dummy variable, coded as 1 if the firm is audited by a big-4 auditor, and 0 otherwise (e.g. Premuroso and Bhattacharya, 2008); foreign assets ratio (FORASSET) represents the percentage of total assets held in a foreign country (e.g. Casterella et al., 2004; Seetharaman et al., 2002); foreign subsidiaries (FORSUB) equals the natural logarithm of 1 plus the number of foreign subsidiaries (e.g. Leventis et al., 2011); RTARATIO is the ratio of total receivables to total assets (e.g. De George et al., 2013); ITARATIO is the ratio of total inventory to total assets (e.g. De George et al., 2013); year dummies represent the dummy variables that reflect the 2005– 2012 period in the US and the 2004–2011 period in Japan; and industry dummies are based on the two-digit US Standard Industry Classification (SIC) and represent 11 industries: agriculture, forestry and fishing; construction; finance, insurance and real estate; mining; public administration; retail trade; services; transportation and communication; electric, gas and sanitary; wholesale trade; manufacturing. 4. Results and discussion 4.1. Descriptive statistics and univariate analysis Table 1 (Panel A) reports the descriptive statistics of the main variables in our regression model. In the US data set, shown in Column (1), the mean (median) for audit fees is 13.85 (13.86), with a range of 11.71 to 16.07. The mean (median) for firm size is 19.16 (19.15), with a range of 16.13 to 22.32. In the Japanese data set, shown in Column (2), the mean (median) for audit fees is 13.09 (12.98), with a range of 11.93 to 14.95. The mean (median) for firm size is 19.38 (19.31), with a range of 16.69 to 22.24. The comparisons of means (t-tests) reported in Column (1) in Table 1 (Panel B) indicate a positive level of statistical significance for audit fees, but a negative level of statistical significance for firm size when comparing the US and Japanese firms. These results are consistent with the Wilcoxon–Mann Whitney test (z-statistic) results presented in Column (2).


95

Table 1 Comparative descriptive statistics and univariate tests.*,† Panel A: Descriptive statistics for key variablea US data set Column (1)

AUDITFEES FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ FORASSET FORSUB RTARATIO ITARATIO

Japan data set Column (2)

Mean

Med

Min

Max

SD

Mean

Med

Min

Max

SD

13.85 19.16 0.56 0.1 0.14 1.38 0.06 1.63 0.11 0.39

13.86 19.15 0.24 0.09 0.09 1.05 0 1 0.08 0.36

11.71 16.13 0 −3.27 −0.28 0.15 −0.06 0.00 −0.09 −0.99

16.07 22.32 2.79 3.42 0.91 4.56 0.43 10 0.91 0.47

1.19 1.71 0.76 1.39 0.28 1.14 0.41 2.06 0.12 0.43

13.09 19.38 0.35 0.27 0.09 0.71 0.05 1.41 0.21 0.49

12.98 19.31 0.14 0.05 0.08 0.57 0.00 0.00 0.2 0.52

11.93 16.69 0 −2.31 −0.18 0.15 0.00 0.00 −0.002 −0.81

14.95 22.24 1.79 5.46 0.46 2.09 0.32 9 0.88 0.26

0.79 1.49 0.48 1.47 0.17 0.48 0.13 1.94 0.15 0.31

Panel B: Comparative univariate tests for key variablesa,b,c,d US versus Japan

AUDITFEES FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ FORASSET FORSUB RTARATIO ITARATIO

Mean-comparison t-test Column (1)

Wilcoxon-Mann Whitney z-test Column (2)

65.12*** −12.8*** 35.25*** −12.35*** 22.11*** 83.75*** 3.49** 12.27*** −71.42*** −29.91***

53.24*** −11.5*** 12.85*** 3.58*** 8.39*** 70.32*** 8.5*** 27.36*** −73.59*** −41***

a AUDITFEES = natural logarithm of audit fees; FIRMSIZE = natural logarithm of value of total assets at the end of fiscal year; DERATIO = long-debt to total equity; ΔEPS = change of EPS, EPSt – EPSt–1; ΔSALES = change of sales ratio, (SALESt – SALESt–1) / SALESt–1; TOBINSQ = Tobin’s Q, market value of stock and book value of debt divided by book value of total assets; FORASSET = foreign assets ratio, the percentage of total assets held in foreign country; FORSUB = foreign subsidiaries, natural logarithm of 1 plus the number of foreign subsidiaries; RTARATIO = ratio of total receivables to total assets; ITARATIO = ratio of total inventory to total assets. b Mean–comparison test (t-test) tests the statistical significance level of difference between two means. c Wilcoxon–Mann Whitney test (z-statistic) is used to examine the statistical significance level of difference between two medians. d † if p < 0.10; * if p < 0.05; ** if p < 0.01; *** if p < 0.001 (two-tailed p-values are used in determining significance).

4.2. Multivariate analysis and tests of differences between the estimated coefficients We use the pooled ordinary least square (OLS) regression estimation technique in this study. To estimate regression coefficients without bias, the following procedures are carried out. First, all continuous variables are winsorized at the 1st and 99th percentiles to moderate the possible effects of extreme outliers. Second, to consider the potential issue of multicollinearity in our regression model, we compute Pearson correlation matrices for the US and Japanese data sets separately. As shown in Table 2 (Panels A and C), there are no correlation coefficients between pairs of independent variables that are greater than the critical value of 0.80. We also compute variance inflation factors (VIFs) to detect potential multicollinearity problems in our regression model because as Gujarati (2003) claims, multicollinearity can exist even if the correlation coefficient between independent variables is low. The results, as reported in Table 2 (Panel B and D), show that the largest VIFs are 1.6 and 2.08 for the US and Japanese data sets, respectively. These values are well below the VIF critical value of 10 (see e.g. Gujarati, 2003), therefore, we conclude that our regression model is free of multicollinearity bias. Columns (1) and (2) of Table 3 provide the regression results for our basic model (i.e. Model 1), which examines the effects of the independent variables on audit fees in the US and Japan. In brief, the results indicate an adjusted R2 of 0.7668 with an F-statistic of 4662.97 for the US data, and an adjusted R2 of 0.6433 with an F-statistic of 1063.04 for the Japanese data. The high adjusted R2 values and F-statistics suggest that the dependent variable (i.e. AUDITFEES) is well explained by the independent variables and control variables in the basic model. The US results reported in Column (1) of Table 3 (Panel A) indicate a statistically significant negative β1 of −0.253 (t = −2.24, p < 0.05) and a statistically significant positive β2 of 0.496 (t = 137.32, p < 0.001). Moreover, the Japanese data results reported in Column (2) of Table 3 (Panel A) show a statistically significant negative β1 of −0.513 (t = −2.52, p < 0.05) and a statistically significant positive β2 of 0.337 (t = 36.36, p < 0.001). The negative coefficient between XBRL and audit fees thus supports H1a. In relation to this particular finding, Rezaee (2009) argues that financial statements in XBRL format facilitate and enhance internal control effectiveness and efficiency and thus reduce relevant costs, including those for verification, substantive tests and audit engagement. The positive coefficient between firm size and audit fees is also consistent with the extant literature

96


Table 2 Collinearity diagnostics.a,b,c Panel A: Pearson correlation matrix for the US data set Variables

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

( 1) XBRL ( 2) FIRMSIZE ( 3) DERATIO ( 4) ΔEPS ( 5) ΔSALES ( 6) TOBINSQ ( 7) BIG4 ( 8) FORASSET ( 9) FORSUB (10) RTARATIO (11) ITARATIO

1.000 0.211*** 0.056*** 0.063*** 0.025*** −0.023** 0.085*** 0.073*** 0.075*** −0.051*** −0.055***

1.000 0.342*** 0.084*** −0.015* −0.088*** 0.511*** 0.363*** 0.299*** 0.022** −0.146***

1.000 0.001 −0.028*** −0.257*** 0.226*** 0.068*** −0.014† −0.126*** −0.494***

1.000 0.268*** 0.146*** 0.049*** 0.043*** 0.038*** 0.039*** −0.005

1.000 0.259*** 0.01 −0.021** −0.011 0.034*** −0.032***

1.000 0.035*** −0.015* 0.045*** −0.06*** 0.14***

1.000 0.15*** 0.111*** −0.116*** −0.171***

1.000 0.074*** 0.23*** 0.034***

1.000 0.204*** 0.124***

1.000 0.394***

1.000

Panel B: VIF diagnostic for the US data set

VIF

XBRL

FIRMSIZE

DERATIO

ΔEPS

ΔSALES

TOBINSQ

BIG4

FORASSET

FORSUB

RTARATIO

ITARATIO

1.05

1.60

1.23

1.04

1.10

1.18

1.40

1.07

1.17

1.19

1.36

Panel C: Pearson correlation matrix for the Japan data set Variables

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

( 1) XBRL ( 2) FIRMSIZE ( 3) DERATIO ( 4) ΔEPS ( 5) ΔSALES ( 6) TOBINSQ ( 7) BIG4 ( 8) FORASSET ( 9) FORSUB (10) RTARATIO (11) ITARATIO

1.000 0.047*** 0.008 −0.044*** 0.076*** −0.302*** 0.007 −0.029*** −0.008 −0.05*** 0.037***

1.000 0.175*** 0.053*** −0.002 −0.082*** 0.199*** 0.409*** 0.466*** 0.15*** 0.057***

1.000 −0.029*** −0.031*** −0.142*** −0.024** 0.05*** 0.057*** −0.166*** −0.395***

1.000 0.344*** 0.156*** 0.028*** 0.032*** 0.041*** 0.038*** 0.014†

1.000 0.146*** 0.047*** −0.017* −0.003 0.016*** 0.012

1.000 0.091*** 0.119*** 0.119*** −0.272*** −0.096***

1.000 0.089*** 0.108*** −0.003 −0.013

1.000 0.073*** 0.065*** 0.053***

1.000 0.043*** 0.019*

1.000 0.061***

1.000

Panel D: VIF diagnostic for the Japan data set

VIF

XBRL

FIRMSIZE

DERATIO

ΔEPS

ΔSALES

TOBINSQ

BIG4

FORASSET

FORSUB

RTARATIO

ITARATIO

1.13

1.46

1.16

1.10

1.12

1.35

1.06

1.78

2.08

1.40

1.27

a AUDITFEES = natural logarithm of audit fees; FIRMSIZE = natural logarithm of value of total assets at the end of fiscal year; DERATIO = long-debt to total equity; ΔEPS = change of EPS, EPSt – EPSt–1; ΔSALES = change of sales ratio, (SALESt – SALESt–1) / SALESt–1; TOBINSQ = Tobin’s Q, market value of stock and book value of debt divided by book value of total assets; BIG4 = Big 4 auditor, coded 1 if the firm is audited by a Big4 auditor, 0 otherwise; FORASSET = foreign assets ratio, the percentage of total assets held in foreign country; FORSUB = foreign subsidiaries, natural logarithm of 1 plus the number of foreign subsidiaries; RTARATIO = ratio of total receivables to total assets; ITARATIO = ratio of total inventory to total assets. b The critical value of the VIF to test for multicollinearity is 10. Gujarati (2003) suggests that there is no evidence of multicollinearity unless the VIF of a variable exceeds 10. All values used in this study were well below this critical level. c † if p < 0.10; * if p < 0.05; ** if p < 0.01; *** if p < 0.001 (two-tailed p-values are used in determining significance).

(see e.g. Firth, 1985; Fleischer and Goettsche, 2012). Our results suggest that compared with small firms, large firms have more complex financial accounts and transactions, higher risks of potential litigation and loss of goodwill (Firth, 1985) and higher political costs because of public visibility (Premuroso and Bhattacharya, 2008). In terms of H1b concerning the joint effect of XBRL and firm size on audit fees, we find that the coefficients reported in Columns (1) and (2) in Table 3 (Panel A) are significantly positive for both the US (β = 0.009, t = 2.59, p < 0.01) and Japanese (β = 0.056, t = 5.5, p < 0.001) data sets, respectively. Nevertheless, although the coefficients are positive, the overall effects of these coefficients are noticeably reduced after the XBRL filing was officially mandated in both the US and Japan. This finding is evidenced by the joint coefficients of XBRL × FIRMSIZE and the coefficients of FIRMSIZE and XBRL (i.e. 0.252 versus 0.496 and −0.253 with a 49.19% decrease for the US data; −0.12 versus 0.337 and −0.513 with a 135.61% decrease for the Japanese data).5 Overall, our results indicate that XBRL has a main negative effect on audit fees that is weaker for larger firms. Thus, our results show that XBRL moderates the association between firm size and audit fees, so H1b is supported. We conjecture that the positive interaction observed may possibly turn negative in the future as the compliance costs are incurred in the short-term (i.e. 2009–2012 for the US and 2008–2011 for Japan) impact on the incremental effect of XBRL and FIRMSIZE. Possible explanations of the positive coefficients identified in our findings could be related to XBRL compliance costs (e.g. XBRL-enabled applications’ set-up and training costs) incurred in the short-term which might be more substantial for larger firms relative to their smaller counterparts (e.g. Bergeron, 2003; DiPiazza and Eccles, 2002; Ramin and Reiman, 2013).

5

The method for computing the joint effects in our regression model is consistent with Wooldridge (2006).


97

Table 3 Primary results: regression results for all firms. Panel A: Regression results for examining H1a and H1b US data seta,b,d Column (1)

Japan data seta,b,d Column (2)

Independent variable

Expected sign

β

t

β

t

Intercept XBRL FIRMSIZE XBRL × FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ BIG4 FORASSET FORSUB RTARATIO ITARATIO Year Effect Industry Effect Adjusted R2 F-statistic Observations

N/A – [H1a] + −/+ [H1b] ? ? ? ? ? ? ? ? ?

4.127 −0.253 0.496 0.009 0.088 −0.006 −0.071 −0.006 0.34 0.058 0.046 0.316 −0.445 Yes Yes 0.7668 4662.97*** 17010

63.43*** −2.24* 137.32*** 2.59** 13.29*** −1.85† −4.29*** −1.37 28.68*** 4.2*** 23.22*** 7.86*** −25.87***

5.736 −0.513 0.337 0.056 0.21 0.035 −0.558 0.055 0.308 0.075 0.004 −0.374 −0.164 Yes Yes 0.6433 1063.04*** 7067

31.28*** −2.52* 36.36*** 5.5*** 15.01*** 7.21*** −15.37*** 3.6*** 20.31*** 1.7† 1.05 −6.82*** −3.98***

Panel B: t-test for differences between the estimated coefficients (US and Japan) for H2 USA versus Japan: coefficient t-testc,d Independent variable

t-difference

Intercept XBRL FIRMSIZE XBRL × FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ BIG4 FORASSET FORSUB RTARATIO ITARATIO

– 1.03 14.54*** −3.7*** −7.18*** −6.54*** 11.11*** −3.43** 1.57 −0.33 9.74*** 9.48*** −5.7***

Notes: This table reports the results of the pooled OLS regression models: AUDITFEESi = α + β1XBRL i + β 2FIRMSIZEi + β 3 XBRL i × FIRMSIZEi + β 4 DERATIOi + β5 ΔEPSi + β6 ΔSALESi + β7 TOBINSQ i 8

11

j =1

k =1

+ β8BIG4i + β9FORASSETi + β10FORSUBi + β11RTARATIOi + β12ITARATIOi + γ j ∑ YEAR i + ηk ∑ INDUSTRYi + ε i

where, AUDITFEES = natural logarithm of audit fees; XBRL = filing financial statement in XBRL format, coded as 1 if company is an XBRL filer, 0 otherwise; FIRMSIZE = natural logarithm of value of total assets at the end of fiscal year; DERATIO = long-debt to total equity; ΔEPS = change of EPS, EPSt – EPSt–1; ΔSALES = change of sales ratio, (SALESt – SALESt–1) / SALESt–1; TOBINSQ = Tobin’s Q, market value of stock and book value of debt divided by book value of total assets; BIG4 = BIG4 = Big 4 auditor, coded 1 if the firm is audited by a Big4 auditor, 0 otherwise; FORASSET = foreign assets ratio, the percentage of total assets held in foreign country; FORSUB = foreign subsidiaries, natural logarithm of 1 plus the number of foreign subsidiaries; RTARATIO = ratio of total receivables to total assets; ITARATIO = ratio of total inventory to total assets. a We conducted the Ramsey’s Regression Specification Error Test (RESET) using powers of the fitted values of dependent variable. The results indicated that the model has no omitted variables. b All continuous variables, except dichotomous variables are winsorized at the 1st and 99th percentile to reduce the possible effects of outliers. c White’s (1980) t-statistic adjusted for heteroskedasticity are used to estimate the significance level of the differences in each pair of estimated coefficients reported each country. The negative and significant t-statistic for the coefficients XBRL and FIRMSIZE would confirm H2: ⎧ fUSA (β1XBRL , β 2FIRMSIZE, β 3 XBRL ∗ FIRMSIZE, β 4 DERATIO, β5 ΔEPS, β6 ΔSALES, ⎪β7 TOBINSQ , β8BIG4, β9 FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO) ⎪ i ,t t -difference (USA-JAPAN) = ⎨ ⎪ f JAPAN (β1XBRL , β 2FIRMSIZE, β 3 XBRL ∗ FIRMSIZE, β 4DERATIO, β5 ΔEPS, β6 ΔSALES, ⎪⎩β7 TOBINSQ , β8BIG4, β9FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO)i,t d

† if p < .10, * if p < .05; ** if p < .01; *** if p < .001. All tests are two-tailed.

98


As far as the control variables for the US data are concerned, we find that DERATIO, BIG4, FORASSET, FORSUB and RTARATIO are positively associated with audit fees, whereas ΔEPS, ΔSALES and ITARATIO exhibit negative associations. However, the other control variable, TOBINSQ, shows no statistical effect. For the Japanese data, we find that DERATIO, ΔEPS, TOBINSQ, BIG4 and FORASSET are positively associated with audit fees, whereas ΔSALES, RTARATIO and ITARATIO show negative associations. The remaining variable, FORSUB, shows no statistical effect. H2 examines whether the effect of XBRL use on audit fees is more significant for publicly listed firms in the US than in Japan. We use the expanded model (i.e. Model 2) and calculate t-statistics to estimate the significance level of the differences in the pairs of β1, USA versus β1, JAPAN and β3, USA versus β3, JAPAN (see e.g. White, 1980). Specifically, we run our regression model including both data sets using White’s (1980) t-statistic for testing the significance level between the two estimated regression coefficients.6 According to the results presented in Panel B of Table 3, there is no significant difference between β1, USA and β1, JAPAN (t = 1.03, p > 0.1). However, we find that there is a significant difference between β3, USA and β3, JAPAN (t = −3.7, p < .001). This means that the t-statistic for the pair-wise difference in the estimated β3s shows that the effect of XBRL × FIRMSIZE for US firms is greater than the effect of XBRL × FIRMSIZE for Japanese firms. Thus, XBRL adoption in both countries led to reduced audit fees, although, overall, the effect of XBRL on audit fees is higher in the US than in Japan. Thus, H2 is supported. These findings are consistent with our argument that US publicly listed firms face a higher threat of litigation (Ginsburg and Hoetker, 2006) and have a more dispersed equity structure and effective corporate governance system than their Japanese counterparts (Fukukawa et al., 2006; Shan and Round, 2012). We also argue that another possible explanation for the greater effect of XBRL on audit fees in the US relative to Japan may be related to the size of the XBRL taxonomies7 used in these jurisdictions. Larger taxonomies can allow greater disclosure and therefore decrease information processing costs for both auditors and investors (Bergeron, 2003). When XBRL filing became mandatory in Japan in 2008, the Japanese GAAP XBRL taxonomy available to Japanese filers comprised approximately 5000 tags (Mitsui and Nakagaito, 2012). In comparison, the US GAAP XBRL taxonomy available to US filers in 2009 comprised approximately 14,000 tags (White, 2012) and this number had grown to approximately 15,000 by 2012 (EY, 2012). Additionally, whereas filers in Japan were mandated to submit primary financial statement data in XBRL format, filers in the US were expected to submit primary financial statement data in the first year of mandatory filing and after the first year this requirement was extended to include all information including footnotes (FSA, 2008; Mitsui and Nakagaito, 2012). This suggests that US filers can disclose more financial information in XBRL format than their Japanese counterparts, which contributes to the greater effect of XBRL on audit fees in the US compared to Japan. 4.3. Robustness checks We carry out several robustness checks. First, we assess the robustness of our primary results by removing financial institutions from our data set. Prior studies of audit fees typically exclude firms operating in the financial services industry (e.g. Felix et al., 2001; Gonthier-Besacier and Schatt, 2007; Thinggaard and Kiertzner, 2008) because their capital structure and performance indicators, such as the debt-to-equity ratio, return on assets, and Tobin’s Q, differ from those of non-financial firms. Accordingly, we remove all of the firms operating in the financial services industry from our sample and analyze the remaining firms. Our sample sizes are reduced to 16,597 and 7005 firm-year observations for the US and Japan, respectively. We test Models 1 and 2 for our robustness check. As shown in Table 4 (Panels A and B), the empirical results are consistent with our primary findings, thus the inclusion of financial firms in our sample does not significantly influence our results. Second, a robustness of our primary results is also considered in terms of the global financial crises (GFC) of 2008 because the crisis brought economic and financial exogenous shocks and such shocks are likely to have had a negative effect on firms’ market performance (see e.g. Aldamen et al., 2012; Kestens et al., 2012). Thus, the long four-year pre- and post-XBRL windows may open up the possibility of confounding factors. For instance, the GFC may have resulted in a sharp increase in systematic risks, which may have led to a significant increase in audit fees. Accordingly, to provide a robustness check on the primary results we use a two-year short window for pre- and post-XBRL (i.e. for the US data, the 2005–2006 and 2011–2012 periods are denoted as pre- and post-XBRL, respectively; for the Japanese data, the 2004–2005 and 2010–2011 periods are denoted as pre- and post-XBRL, respectively). The (untabulated) results are consistent with our primary findings.8 5. Conclusion The objective of this study was to analyze the effect of XBRL on audit fees, and consider whether and how this effect varies across different jurisdictions, namely, the US and Japan. We find that XBRL use is inversely associated with audit fees, which are,

6

Aharony et al. (2010) apply White’s (1980) t-statistic method in a similar way in their study. More specifically, a taxonomy is a data dictionary that maps XBRL tags onto financial accounting concepts, while also defining their relationships and processing rules (Troshani and Lymer, 2010). Given that XBRL taxonomies are developed on a jurisdictional basis, the taxonomy of a jurisdiction reflects both its accounting standards and its GAAP. 8 We also use a three-year window (i.e. for the US data, the time spans of 2005–2007 and 2010–2012 are denoted as pre- and post-XBRL, respectively; for the Japanese data, the time spans of 2004–2006 and 2009–2011 are denoted as pre- and post-XBRL, respectively) to provide an additional robustness check of the primary results. The (untabulated) results are also consistent with the primary findings. 7


99

Table 4 Robustness check: regression results for non-financial firms. Panel A: Regression results for examining H1a and H1b US data seta,b,d Column (1)

Japan data seta,b,d Column (2)

Independent variable

Expected sign

β

t

β

T

Intercept XBRL FIRMSIZE XBRL × FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ BIG4 FORASSET FORSUB RTARATIO ITARATIO Year effect Industry effect Adjusted R2 F-statistic Observations

N/A – + −/+ ? ? ? ? ? ? ? ? ?

4.24 −0.286 0.494 0.011 0.085 −0.004 −0.072 −0.006 0.343 0.057 0.046 0.326 −0.448 Yes Yes 0.7652 4507.11*** 16597

62.95*** −2.49** 135.28*** 1.84† 12.56*** −1.91† −4.27*** −1.33 28.42*** 4.13*** 23.13*** 7.91*** −25.57***

5.734 −0.523 0.337 0.057 0.217 0.037 −0.556 0.052 0.307 0.075 0.003 −0.413 −0.14 Yes Yes 0.6447 1057.15*** 7005

31.15*** −2.56* 36.25*** 5.51*** 15.34*** 7.49*** −15.2*** 3.4** 20.14*** 1.7† 0.83 −7.41*** −3.33**

Panel B: test for differences between the estimated coefficients (US and Japan) for H2 USA versus Japan: coefficient t-testc,d Independent variable

t-difference

XBRL FIRMSIZE XBRL × FIRMSIZE DERATIO ΔEPS ΔSALES TOBINSQ BIG4 FORASSET FORSUB RTARATIO ITARATIO

0.93 14.35*** −3.58*** −7.7*** −6.79*** 10.94*** −3.26** 1.71† −0.36 9.93*** 9.97*** −6.12***

Notes: This table reports the results of the pooled OLS regression models: AUDITFEESi = α + β1XBRL i + β 2FIRMSIZEi + β 3 XBRL i × FIRMSIZEi + β 4 DERATIOi + β5 ΔEPSi + β6 ΔSALESi + β7 TOBINSQ i 8

11

j =1

k =1

+ β8BIG4i + β9FORASSETi + β10FORSUBi + β11RTARATIOi + β12ITARATIOi + γ j ∑ YEAR i + ηk ∑ INDUSTRYi + ε i

where, AUDITFEES = natural logarithm of audit fees; XBRL = filing financial statement in XBRL format, coded as 1 if company is an XBRL filer, 0 otherwise; FIRMSIZE = natural logarithm of value of total assets at the end of fiscal year; DERATIO = long-debt to total equity; ΔEPS = change of EPS, EPSt – EPSt–1; ΔSALES = change of sales ratio, (SALESt – SALESt–1) / SALESt–1; TOBINSQ = Tobin’s Q, market value of stock and book value of debt divided by book value of total assets; BIG4 = BIG4 = Big 4 auditor, coded 1 if the firm is audited by a Big4 auditor, 0 otherwise; FORASSET = foreign assets ratio, the percentage of total assets held in foreign country; FORSUB = foreign subsidiaries, natural logarithm of 1 plus the number of foreign subsidiaries; RTARATIO = ratio of total receivables to total assets; ITARATIO = ratio of total inventory to total assets. a We conducted the Ramsey’s Regression Specification Error Test (RESET) using powers of the fitted values of dependent variable. The results indicated that the model has no omitted variables. b All continuous variables, except dichotomous variables are winsorized at the 1st and 99th percentile to reduce the possible effects of outliers. c White’s (1980) t-statistic adjusted for heteroskedasticity are used to estimate the significance level of the differences in each pair of estimated coefficients reported each country. The negative and significant t-statistic for the coefficients XBRL and FIRMSIZE would confirm H2: ⎧ fUSA (β1XBRL , β 2FIRMSIZE, β 3 XBRL ∗ FIRMSIZE, β 4 DERATIO, β5 ΔEPS, β6 ΔSALES, ⎪β7 TOBINSQ , β8BIG4, β9 FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO) ⎪ i ,t t -difference (USA-JAPAN) = ⎨ ⎪ f JAPAN (β1XBRL , β 2FIRMSIZE, β 3 XBRL ∗ FIRMSIZE, β 4DERATIO, β5 ΔEPS, β6 ΔSALES, ⎪⎩β7 TOBINSQ , β8BIG4, β9FORASSET, β10FORSUB, β11RTARATIO, β12ITARATIO)i,t d

† if p < .10, * if p < .05; ** if p < .01; *** if p < .001. All tests are two-tailed.

100


in turn, positively associated with firm size. We also find evidence that, overall, XBRL moderates the association between firm size and audit fees in both the US and Japan. However, these moderation effects are weaker among Japanese firms. Although XBRL is expected to bring significant benefits for financial reporting internationally, there are increasing calls in the literature from both academics and practitioners for empirical evidence in response to the increasing rhetoric about its espoused benefits (see e.g. Liu et al., 2014; Premuroso and Bhattacharya, 2008; Roohani et al., 2009). Our study represents a direct response to these calls and thus contributes to the extant literature. In particular, our study appears to be the first to examine the effect of XBRL on audit fees for publicly listed firms across two major countries: the US and Japan. Additionally, given the significant costs associated with introducing and implementing XBRL for digital business reporting, our study is also one of the first to provide empirical evidence helpful to the on-going XBRL viability debate, specifically showing that its economic benefits are materializing in practice and thereby supporting the business case of XBRL use in the regulatory domain (Bai et al., 2013; Debreceny et al., 2010; Li et al., 2012; Troshani and Lymer, 2010). Our findings have important implications for all stakeholders of the business information supply chain, including preparers and users of financial statements, auditors and regulators. In particular, they can inform policy-making and practice for auditing, while offering evidence that can inform the development of XBRL diffusion strategies in countries where XBRL use is currently being considered. Our findings are also likely to assist managers in reconsidering the allocation of scarce internal auditing resources to other financial reporting activities after XBRL is introduced (Helliar et al., 1996). Finally, our findings have broader implications as increased financial reporting transparency through XBRL use can improve the credibility of firms and mitigate information asymmetry between management and shareholders, potentially resulting in an overall improvement of investors’ confidence. This study has several limitations. First, it examines the association between audit fees and XBRL without considering other potential effects such as corporate governance. The literature suggests that corporate governance factors can also influence financial reporting transparency (e.g. Demsetz and Lehn, 1985; Denis and McConnell, 2003; Fama and Jensen, 1983; La Porta et al., 1999) and audit quality (e.g. Lin and Liu, 2009a, 2009b; Shan, 2014). Future research could investigate how XBRL use, whether in mandatory or voluntary settings, interacts with corporate governance variables and their joint effect on transparency and auditing costs. Second, as our study does not consider how auditing processes are influenced by XBRL, future research could explore how XBRL affects the scope of the traditional audit function and the manner in which internal control activities are being carried out. Future research could also examine the extent to which XBRL contributes to achieving the uniformity and convergence of auditing practices internationally, especially given the on-going efforts of the International Auditing and Assurance Standards Board, a standing committee of the International Federation of Accountants, to enhance the uniformity of auditing practices and related services worldwide (Gupta, 2004). Acknowledgements We are grateful for the constructive comments of the anonymous reviewer and the co-editor, Professor Bin Srinidhi. We also acknowledge the constructive comments received from the workshop participants at University of Adelaide and University of Western Sydney. References Abbott, L.J., Parker, S., Peters, G.F., 2006. Earnings management, litigation risk, and asymmetric audit fee responses. Auditing 25 (1), 85–98. Abbott, L.J., Parker, S., Peters, G.F., Rama, D.V., 2007. Corporate governance, audit quality, and the Sarbanes-Oxley Act: evidence from internal audit outsourcing. Account. Rev. 82 (4), 803–835. Abdolmohammadi, M., Harris, J., Smith, K., 2002. Government financial reporting on the Internet: the potential revolutionary effects of XBRL. J. Gov. Financ. Manage. 51 (2), 24–26, 28–31. Aharony, J., Barniv, R., Falk, H., 2010. The impact of mandatory IFRS adoption on equity valuation of accounting numbers for security investors in the EU. Eur. Account. Rev. 19 (3), 535–578. Aldamen, H., Duncan, K., Kelly, S., McNamara, R., Nagel, S., 2012. Audit committee characteristics and firm performance during the global financial crisis. Account. Finance 52, 971–1000. Alleyne, P., Hudaib, M., Pike, R., 2013. Towards a conceptual model of whistle-blowing intentions among external auditors. Br. Account. Rev. 45 (1), 10–23. Bai, Z., Sakaue, M., Takeda, F., 2013. The impact of XBRL adoption on the information environment in Japan. Paper presented at the EAA: 36th Annual Congress. (accessed 05.06.13.). Beatty, R., 1993. The economic determinants of auditor compensation in the initial public offering market. J. Account. Res. 31 (2), 294–302. Bedard, J.C., Hoitash, R., Hoitash, U., 2009. Evidence from the United States on the effect of auditor involvement in assessing internal control over financial reporting. Int. J. Account. 13 (2), 105–125. Bergeron, B., 2003. Essentials of XBRL: Financial Reporting in the 21st Century. John Wiley & Sons, Hoboken, NJ. Bizarro, P.A., Garcia, A., 2010. XBRL-beyond basics. CPA J. 80 (5), 62–71. (accessed 04.06.14.). Blankenspoor, E., Miller, B.P., White, H.D., 2014. Initial evidence on the market impact of the XBRL mandate. Rev. Account. Stud. 20 (1), doi:10.1007/s1114211013-19273-11144. Boo, E., Sharma, D., 2008. Effect of regulatory oversight on the association between internal governance characteristics and audit fees. Account. Finance 48 (1), 51–71. Boritz, J.E., No, W.G., 2008. The SEC’s XBRL voluntary filing program on EDGAR: a case for quality assurance. Curr. Issues Auditing 2 (2), A36–A50. Boyle, D.M., Boyle, J.F., Carpenter, B.W., 2014. The SEC’s renewed focus on accounting fraud. CPA J. 84 (2), 68–72. (accessed 04.06.14.). Brands, K., 2013. XBRL and the audit data standards. Institute of Management Accountants. (accessed 04.06.14.). Brown, D.L., Shu, S.Z., Soo, B.S., Trompeter, G.M., 2013. The insurance hypothesis: an examination of KPMG’s audit clients around the investigation and settlement of the tax shelter case. Auditing 32 (4), 1–24. Casterella, J.R., Francis, J.R., Lewis, B.L., Walker, P.L., 2004. Auditor industry specialization, client bargaining power, and audit pricing. Auditing 23 (1), 123–140.


101

Chen, J.J., Zhang, H., 2010. The impact of regulatory enforcement and audit upon IFRS compliance – evidence from China. Eur. Account. Rev. 19 (4), 665– 692. Christie, A.A., Zimmermann, J., 1994. Efficient and opportunistic choices of accounting procedures: corporate control contest. Account. Rev. 69 (4), 539–566. Cohen, E.E., 2007. The Need for and Issue Surrounding the Seamless Audit Trail. Working Paper: [email protected]. De George, E.T., Ferguson, C.B., Spear, N.A., 2013. How much does IFRS cost? IFRS adoption and audit fees. Account. Rev. 88 (2), 429–462. DeAngelo, L., 1981. Auditor size and audit quality. J. Account. Econ. 3 (3), 183–199. Debreceny, R., Farewell, S., Piechocki, M., Felden, C., Gräning, A., 2010. Does it add up? Early evidence on the data quality of XBRL filings to the SEC. J. Account. Public Policy 29 (3), 296–306. Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: causes and consequences. J. Polit. Econ. 93, 1155–1177. Denis, D.K., McConnell, J.J., 2003. International corporate governance. J. Financ. Quant. Anal. 38 (1), 1–36. DiPiazza, S.A.J., Eccles, R.G., 2002. Building Public Trust: The Future of Corporate Reporting. John Wiley & Sons, New York. Doolin, B., Troshani, I., 2004. XBRL: a research note. Qual. Res. Account. Manage. 1 (2), 93–104. Doolin, B., Troshani, I., 2007. Organizational adoption of XBRL. Electron. Mark. 17 (3), 199–209. Dopuch, N., 1980. The demand for quality-differentiated audit services in an agency cost setting: an empirical investigation. Paper presented at the Auditing Research Symposium. Dopuch, N., Simunic, D., 1982. The nature of competition in the auditing profession: a descriptive and normative view. In: Buckley, J.W., Weston, J.F. (Eds.), Regulation and the Accounting Profession. Lifetime Learning, pp. 77–94. Du, H., Roohani, S., 2007. Meeting challenges and expectations of continuous auditing in the context of independent audits of financial statements. Int. J. Account. 11 (2), 133–146. Eaglesham, J., 2013. Accounting fraud targeted. Wall St. J. 261 (124), C1–C6. ; 27 May 2013 (accessed 05.05.13.). Eccles, R.G., Krzus, M.P., 2010. One Report: Integrated Reporting for a Sustainable Strategy. Wiley, Hoboken, NJ. Efendi, J., Park, J.D., Subramaniaum, C., 2009. Do XBRL reports have incremental information content? – an empirical analysis. . Efendi, J., Park, J.D., Smith, L.M., 2014. Do XBRL filings enhance informational efficiency? Early evidence from post-earnings announcement drift. J. Bus. Res. 23 (6), 1099–1105. Elder, R., Zhang, Y., Zhou, J., Zhou, N., 2009. Internal control weaknesses and client risk management. J. Account. Auditing Finance 24 (4), 543–579. EY, 2012. Technical line financial reporting development: using the 2012 XBRL US GAAP taxonomy. No. 2012-15. Ernst & Young. (accessed 04.06.13.). Fama, E.F., Jensen, M.C., 1983. The separation of ownership and control. J. Law Econ. 26 (2), 301–325. Felix, J.W.L., Gramling, A.A., Maletta, M.J., 2001. The contribution of internal audit as a determinant of external audit fees and factors influencing this contribution. J. Account. Res. 39 (3), 513–534. Firth, M., 1985. An analysis of audit fees and their determinants in New Zealand. Auditing 4 (2), 23–37. Fleischer, R., Goettsche, M., 2012. Size effects and audit pricing: evidence from Germany. J. Int. Account. Auditing Taxation 21 (2), 156–168. Francis, J.R., 1984. The effect of audit firm size on audit prices: a study of the Australian market. J. Account. Econ. 6 (2), 133–151. Francis, J.R., 2004. What do we know about audit quality? Br. Account. Rev. 36 (4), 345–368. Francis, J.R., Stokes, D., 1986. Audit prices, product differentiation, and scale economies: further evidence from the Australian market. J. Account. Res. 24 (2), 383–393. Francis, J.R., Wang, D., 2004. Investor protection and auditor conservatism: are Big 4 auditors conservative only in the United States? Working Paper: University of Missouri and University of Nebraska. Francis, J.R., Wilson, E.R., 1988. Auditor changes: a joint test of theories relating to agency costs and auditor differentiation. Account. Rev. 63 (4), 663–682. Frantz, P., 1999. Auditor’s skill, auditing standards, litigation, and audit quality. Br. Account. Rev. 31 (2), 151–183. FSA, 2008. FSA launches new electronic corporate disclosure system (EDINET). 17 March 2008. Financial Services Authority (FSA). (accessed 03.02.13.). Fukukawa, H., Mock, T.J., Wright, A., 2006. Audit programs and audit risk: a study of Japanese practice. Int. J. Auditing 10 (1), 41–65. Fung, S.Y.K., Gul, F.A., Krishnan, J., 2012. City-level auditor industry specialization, economies of scale, and audit pricing. Account. Rev. 87 (4), 1281–1307. Ginsburg, T., Hoetker, G., 2006. The unreluctant litigant? An empirical analysis of Japan’s turn to litigation. J. Legal Stud. 35 (1), 31–59. Gonthier-Besacier, N., Schatt, A., 2007. Determinants of audit fees for French quoted firms. Managerial Auditing J. 22 (2), 139–160. Gujarati, D.N., 2003. Basic Econometrics, 4th ed. McGraw-Hill, New York. Gul, F.A., 1999. Audit prices, product differentiation and economic equilibrium. Auditing 18 (1), 90–100. Gul, F.A., Tsui, J.S.L., 1998. A test of the free cash flow and debt monitoring hypothesis: evidence from audit pricing. J. Account. Econ. 24 (2), 219–237. Gul, F.A., Tsui, J.S.L., 2001. Free cash flow, debt monitoring, and audit pricing: further evidence on the role of director equity ownership. Auditing 20 (1), 72–84. Gul, F.A., Jaggi, B.L., Krishnan, G.V., 2007. Auditor independence: evidence on the joint effects of auditor tenure and nonaudit fees. Auditing 26 (2), 117–142. Gupta, K., 2004. Contemporary Accounting, 6th ed. McGraw-Hill, New York. Helliar, C., Lyon, B., Monroe, G.S., Ng, J., Woodliff, D.R., 1996. UK auditors’ perceptions of inherent risk. Br. Account. Rev. 28 (1), 45–72. Hodge, F., Pronk, M., 2006. The impact of expertise and investment familiarity on investors’ use of online financial reporting. J. Account. Auditing Finance 21 (3), 267–292. Hodge, F., Kennedy, J.J., Maines, L.A., 2004. Does search-facilitating technology improve the transparency of financial reporting? Account. Rev. 79 (3), 687–703. Holzinger, A., 2013. New Robocop Tasked with Busting Fraudulent Financial Report Filers. The Institute of Internal Auditors, Financial Services Auditors, Florida. (accessed 04.06.14.). Houque, M.N., Monem, R.M., Zijl, T.V., 2012. Government quality, auditor choice and adoption IFRS: a cross country analysis. Adv. Account. 28 (2), 307– 316. Hwang, J.S., Leem, C.S., Moon, H.J., 2008. A study on relationships among accounting transparency, accounting information transparency, and XBRL. Proceedings of the 2008 Third International Conference on Convergence and Hybrid Information Technology. (accessed 03.02.13.). Janvrin, D., Caster, P., Elder, R., 2010. Enforcement release evidence on the audit confirmation process: implications for standard setters. Res. Account. Regul. 22 (1), 1–17. Jensen, M., Meckling, W., 1976. Theory of the firm: management behavior, agency costs and ownership structure. J. Financ. Econ. 3 (4), 305–360. Kestens, K., van Cauwenberge, P., Bauwhede, H.V., 2012. Trade credit and company performance during the 2008 financial crisis. Account. Finance 52, 1125–1151. Khadaroo, I., 2005. Corporate reporting on the internet: some implications for the audit profession. Managerial Auditing J. 20 (6), 578–591. Khurana, I.K., Raman, K.K., 2004. Litigation risk and the financial reporting credibility of big 4 versus non-big 4 audits: evidence from Anglo-American countries. Account. Rev. 79 (2), 473–495. Kobayashi, T., 2008. Why was the XBRL introduction in Japan successful? Data Interactive, Hitachi. (accessed 03.02.13.). La Porta, R., Lopez-de-Silanes, F., Schleifer, A., 1999. Corporate ownership around the world. J. Finance 54 (2), 471–517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2003. What works in securities laws? Tuck School of Business Working Paper No. 03-22; AFA 2005 Philadelphia Meetings. or http://dx.doi.org/10.2139/ssrn.425880:> (accessed 03.02.13.).

102


La Rosa, F., Caserio, C., 2013. Are auditors interested in XBRL? A qualitative survey of big auditing firms in Italy. In: Mancini, D., Vaassen, E.H.J., Dameri, R.P. (Eds.), Accounting Information Systems for Decision Making. Springer, New York, pp. 13–45. Lai, K.W., 2009. Audit opinion and disclosure of audit fees. J. Account. Auditing Finance 24 (1), 91–114. Leventis, S., Weetman, P., Caramanis, C., 2011. Agency costs and product market competition: the case of audit pricing in Greece. Br. Account. Rev. 43 (2), 112–119. Li, O.Z., Lin, Y., Ni, C., 2012. Does XBRL adoption reduce the cost of equity capital? Social Science Research Network (SSRN) (August 17, 2012). (accessed 03.02.13.). Li, Y., Roge, J.N., Rydl, L., Hughes, J., 2007. Achieving Sarbanes-Oxley compliance with XBRL-based ERP and continuous auditing. Issues Inf. Syst. 8 (2), 430–436. Lin, Z.J., Liu, M., 2009a. The impact of corporate governance on auditor choice: evidence from China. J. Int. Account, Audic. Tax. 18 (1), 44– 59. Lin, Z.J., Liu, M., 2009b. The determinants of auditor switching from the perspective of corporate governance in China. Corp. Govern.: An Int. Rev. 17 (4), 476–491. Liu, C., Wang, T., Yao, L.J., 2014. XBRL’s impact on analyst forecast behavior: an empirical study. J. Account. Public Policy 33 (1), 69–82. Lymer, A., Debreceny, R., 2003. The auditor and corporate reporting on the internet: challenges and institutional responses. Int. J. Auditing 7 (2), 103–120. Lyon, J.D., Maher, M.W., 2005. The importance of business risk in setting audit fees: evidence from cases of client misconduct. J. Account. Res. 43 (1), 133–151. Markus, H.R., Kitayama, S., 1991. Culture and the self: implications for cognition, emotion, and motivation. Psychol. Rev. 98 (2), 224–253. Marston, C., Polei, A., 2005. Corporate reporting on the internet by German companies. Int. J. Account. Inf. Syst. 5 (3), 285–311. Meek, G.K., Roberts, C.B., Gray, S.J., 1995. Factors influencing voluntary annual report disclosures by US, UK and continental European multinational corporations. J. Int. Bus. Stud. 26 (3), 555–572. Mitsui, C., Nakagaito, M., 2012. A New Era of Corporate Disclosure – Will 2013 Upgrade to EDINET and New-Generation XBRL Improve Corporate Valuation Process? vol. 145. Iakyara. Nomura Research Institute, Ltd, Tokyo. (accessed 05.02.13.). Nikkinen, J., Sahlström, P., 2004. Does agency theory provide a general framework for audit pricing? Int. J. Auditing 8, 253–262. Numata, S., Takeda, F., 2010. Stock market reactions to audit failure in Japan: the case of Kanebo and ChuoAoyama. Int. J. Account. 45 (2), 17–199. O’Sullivan, N., 2000. The impact of board composition and ownership on audit quality: evidence from large UK companies. Br. Account. Rev. 32 (4), 397–414. Pratt, J., Stice, J.D., 1994. The effects of client characteristics on auditor litigation risk judgements, required audit evidence, and recommended audit fees. Account. Rev. 69 (4), 639–656. Premuroso, R.F., Bhattacharya, S., 2008. Do early and voluntary filers of financial information in XBRL format signal superior corporate governance and operating performance? Int. J. Account. Inf. Syst. 9 (1), 1–20. PWC, 2011. XBRL reporting risk and the role of internal audit. PricewaterhouseCoopers LLP. (accessed 06.02.13.). Ramin, K., Reiman, C., 2013. IFRS and XBRL: How to Improve Business Reporting through Technology and Object Tracking. Wiley, Cornwall, UK. Rediker, K.J., Seth, A., 1995. Boards of directors and substitution effects of alternative governance mechanisms. Strateg. Manage. J. 16 (2), 85–99. Reynolds, J.K., Francis, J.R., 2001. Does size matter? The influence of large clients on office-level auditor reporting decisions. J. Account. Econ. 30 (3), 375–400. Rezaee, Z., 2009. Corporate Governance and Ethics. John Wiley & Sons, Hoboken, NJ. Roohani, S., Furusho, Y., Koizumi, M., 2009. XBRL: improving transparency and monitoring functions of corporate governance. Int. J. Disclosure Governance 6 (4), 355–369. Roselli, L., 2013. Challenges with XBRL. Financial Executives Research Foundation (FERF), Danvers, MA. (accessed 05.06.14.). Roth, D., 2009. Road map for financial recovery: radical transparency now! Wired Mag. 17 (3), (accessed 05.06.14.). Sarens, G., Abdolmohammadi, M.J., 2011. Monitoring effects of the internal control audit function: agency theory versus other explanator variables. Int. J. Auditing 15 (1), 1–20. Sarens, G., De Beelde, I., Everaert, P., 2009. Internal audit: a comfort provider to the audit committee. Br. Account. Rev. 41 (2), 90–106. Scapens, R.W., 1985. Management Accounting: A Review of Recent Developments. Macmillan Press Ltd, London. Seetharaman, A., Gul, F.A., Lynn, S., 2002. Litigation risk and audit fees: evidence from UK firms listed on US markets. J. Account. Econ. 33 (1), 91–115. SEC, 2009. Interactive data to improve financial reporting: final rule. RIN 3235-AJ71. US Securities and Exchange Commission. (accessed 05.01.13.). Shan, Y.G., 2014. The impact of internal governance mechanisms on audit quality: a study of large listed companies in China. Int. J. Account. Audit. Perform. Eval. 10 (1), 68–90. Shan, Y.G., Round, D., 2012. China’s corporate governance: emerging issues and problems. Mod. Asian Stud. 46 (5), 1316–1344. Simkin, M.G., Rose, J.M., Norman, C.S., 2012. Core Concepts of Accounting Information Systems, 12th ed. John Wiley & Sons, Hoboken, NJ. Simunic, D.A., 1980. The pricing of audit services: theory and evidence. J. Account. Res. 18 (1), 161–190. Simunic, D.A., Stein, M.T., 1996. The impact of litigation risk on audit pricing: a review of the economics and the evidence. Auditing 15, 119–134. Skinner, D.J., Srinivasan, S., 2012. Audit quality and auditor reputation: evidence from Japan. Account. Rev. 87 (5), 1737–1765. Srivastava, R.P., Kogan, A., 2010. Assurance on XBRL instance document: a conceptual framework of assertions. Int. J. Account. Inf. Syst. 11 (3), 261–273. Starr, M., 2012. SEC interview: inside information. Interactive Bus. Rep. 2 (1), 17–22. (accessed 05.02.13.). Taylor, M.H., Simon, D.T., 1999. Determinants of audit fees: the importance of litigation, disclosure, and regulatory burdens in audit engagements in 20 countries. Int. J. Account. 34 (3), 375–388. TheCorporateCouncil, 2011. EU proposes audit reform. The Corporate Council. (accessed 14.12.14.). Thinggaard, F., Kiertzner, L., 2008. Determinants of audit fees: evidence from a small capital market with a joint audit requirement. Int. J. Auditing 12 (2), 141–158. Troshani, I., Doolin, B., 2007. Innovation diffusion: a stakeholder and social network view. Eur. J. Innov. Manage. 10 (2), 176–200. Troshani, I., Lymer, A., 2010. Translation in XBRL standardization. Inf. Technol. People 23 (2), 136–164. Troshani, I., Lymer, A., 2011. Institutionalizing XBRL in the UK: an organizing vision perspective. Proceedings of the 19th European Conference on Information Systems (ECIS2011), Helsinki, 9–11 June. Troshani, I., Rao, S., 2007. Drivers and inhibitors to XBRL adoption: a qualitative approach to build a theory in under-researched areas. Int. J. E-Bus. Res 3 (4), 98–111. Troshani, I., Parker, L.D., Lymer, A., 2014. Institutionalising XBRL for financial reporting: resorting to regulation. Account. Bus. Res. 1–33. http:// dx.doi.org/10.1080/00014788.2014.980772 (accessed 14.12.14.). Tysiac, K., 2013. What accounting fraud risk factors will attract SEC’s attention? J. Accountancy 216 (6), 1–4. (accessed 04.06.14.). Uehara, K., Yamase, M., Miura, S., Tsumakawa, W., Funaki, K., Takaura, K., et al., 2008. J-SOX challenge: efforts to comply with the new Japanese regulation. Inf. Syst. Control J. 5, 1–5. (accessed 05.01.13.). Venkataraman, R., Weber, J.P., Willenborg, M., 2008. Litigation risk, audit quality, and audit fees: evidence from initial public offerings. Account. Rev. 83 (5), 1315–1345. Wallace, W.A., 1984. Internal auditors can cut outside CPA costs. Harv. Bus. Rev. 62 (2), 16–20.


103

White, C.E., 2012. The Accountant’s Guide to XBRL. University of Delaware. White, H., 1980. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 48 (4), 817–838. Wooldridge, J.M., 2006. Introductory Econometrics: A Modern Approach, 3rd ed. South-Western College Publishing, Mason, OH. Yoon, H., Zo, H., Ciganek, A.P., 2011. Does XBRL adoption reduce information asymmetry? J. Bus. Res. 64 (2), 157–163. Zaman, M., Hudaib, M., Haniffa, R., 2011. Corporate governance quality, audit fees and non-audit services fees. J. Bus. Finance Account. 38 (1/2), 165–197.