Web-Based Disclosure About Value Creation Processes: A Monitoring Perspective
We acknowledge financial support from the Social Sciences and Humanities Research Council of Canada and Fonds Québecois de recherche sur la société et la culture (FQRSC), l'Autorité des marchés financiers (Québec), PriceWaterhouseCoopers and KPMG.
Abstract
Adopting a monitoring perspective, this study aims to explain how and why firms provide web-based disclosure about their value creation and its underlying processes. We rely on the balanced scoreboard approach to measure disclosure. Our results suggest that costs incurred by capital markets' participants as well as monitoring by the board and the media drive disclosure. Moreover, we argue and document that a firm's disclosure is actually a part of its governance configuration and influences some board processes.
This paper addresses two complementary research questions. First, to what extent do monitoring considerations drive a firm's web-based disclosure about value creation processes? For that purpose, monitoring refers to financial market monitoring, governance monitoring and media monitoring. Second, how do disclosure, board effectiveness and media coverage interact with one another? In that context, we argue that disclosure represents a facet of a firm's governance configuration.
Corporate managers view strategy formulation and implementation in a multi-dimensional way, focusing on different performance or value creation metrics (e.g., Simons, 2000). Therefore, effective monitoring of a firm's management decisions and actions implies that boards of directors, financial markets participants and the media get access to a comprehensive information set that reflects such underlying value creation initiatives and actions. In that regard, we argue that a firm's disclosure complements its governance and monitoring mechanisms. The mapping between non-financial measures and value creation, as well as the importance of these measures, has been widely recognized by financial analysts, especially in industries where there are sizable intangibles assets (e.g., Dempsey et al., 1997; Healy et al., 1999). Moreover, the management accounting literature is replete with findings showing that balanced scorecard-based metrics are reflective of a firm's future financial performance.
The challenge for an organization is to disclose information about its value creation in a cost-efficient fashion while maximizing its reach. Most organizations typically disclose non-financial information through traditional media vehicles (e.g., annual report) or intermediaries (e.g., press releases to be picked up by the media). However, the advent of the World Wide Web (Web) brings firms to reconsider their disclosure strategies as it allows for direct communications with—current and potential—stockholders, irrespective of their location and without the need for intermediaries. Moreover, there is essentially no marginal distribution cost if additional information is conveyed. Such a context implies that the stewardship relation between a firm's management and its stockholders becomes more direct, dynamic and interactive.
In that regard, the balanced scorecard literature provides a useful template to capture the various dimensions of value creation processes, which derive from successful management of a firm's financial resources, customer relations, internal processes and human capital (e.g., Kaplan and Norton, 2004). In this paper, we rely on a comprehensive measure of disclosure, which emphasizes content, an approach that is consistent with prior work on financial/non-financial disclosure (e.g., Gibbins et al., 1990; Lang and Lundholm, 1993, 1996; Botosan, 1997; Healy et al., 1999; AIMR, 2002). Our disclosure measure comprises only information that is on a firm's website and in an HTML format. It excludes mandated corporate documents which are linked to the website (e.g., audited financial statements in PDF). The decision to disclose on the Web is not mandated by regulatory agencies. Moreover, while regulators have oversight responsibilities over a firm's disclosure activities, their oversight is bound to be less tight for disclosures when there is no specified form or content to rely upon.
In addition, we put forward the view that disclosure is an important facet in its own right of a firm's governance configuration, as it facilitates monitoring by external parties of value creation processes and outcomes. On the one hand, we argue that a firm's propensity to provide information to capital markets is conditioned by (a) the extent of monitoring costs that investors face; (b) its governance structure, particularly at the board level; and (c) media coverage, reflecting its need to defend and legitimize its purpose and the value of its activities to a broader community. On the other hand, we expect the board to rely on disclosure to enhance monitoring of managers and to be concerned as to how the firm is perceived by the media.
Our sample comprises Canada's largest publicly traded firms representing close to 80% of its total stock market capitalization. There are some advantages in using Canadian data. First, regulators, corporations and investors have been relying on the Web as a disclosure platform for many years now. For instance, since 1997, all corporate documents for publicly traded firms must be posted on the SEDAR website, which is administered by Canadian securities regulators. Second, while Canadian capital markets evolve within a legal and regulatory regime that is similar to U.S. ones, Canadian firms exhibit governance characteristics that are closer to European and Asian firms, notably in terms of ownership and board structures (Morck et al., 2000; Roe, 2003).
Results are consistent with financial market, board and media monitoring driving web-based disclosure about value creation. This paper extends prior research in several ways. First, there are extensive literatures on the determinants of financial performance disclosure (e.g., Healy and Palepu, 2001) and environmental disclosure (e.g., Berthelot et al., 2003; Aerts and Cormier, 2009), either mandated or voluntary. However, from governance and strategic perspectives, managers must consider many other dimensions of organizational performance, the balanced scorecard being one tool to achieve that objective. In that regard, there is limited evidence on the broad-based comprehensive voluntary disclosure that firms provide with respect to the many aspects of their internal performance management. Second, early work on Internet reporting focuses typically on financial statements (e.g., Ashbaugh et al., 1999). More recently, there have been some attempts to investigate corporate social responsibility disclosure (e.g., Patten, 2002; Unerman and Bennett, 2004; Cormier et al., 2009a). We consider that Internet reporting encompasses other dimensions such as intellectual capital. Third, our disclosure measure takes into account all relevant information available on a firm's website, not only its investor relations web pages (e.g., Bollen et al., 2006). Finally, we argue that a firm's governance and disclosure practices are closely intertwined, with governance affecting disclosure and disclosure being a governance mechanism (e.g., Markarian et al., 2007).
Our study extends prior work by Cormier et al. (2009b) on several dimensions. First, in contrast to Cormier et al. (2009b) who adopt an uni-dimensional perspective, the current paper argues that monitoring considerations, either by internal (e.g., board) or external (e.g., market participants) parties, underlies a firm's disclosure profile, thus providing a more comprehensive and integrated view of disclosure determination. Second, we put forward the view that disclosure is an important facet in its own right of a firm's governance configuration. Third, we explicitly take into consideration the interrelations between disclosure, governance and the media. More specifically, directors' monitoring role seems to be influenced by media exposure. This suggests that directors perceive themselves to be accountable to a wider audience than just investors. Finally, by controlling for lagged disclosure, we are cognizant of the fact that disclosure evolves incrementally over time.
CONCEPTUAL BACKGROUND
Value Creation for Shareholders
According to Kaplan and Norton (1996, 2004), the creation of long-term shareholder value encompasses four complementary phases that integrate financial and non-financial information: learning and growth, internal processes, customer and financial. Within each phase, specific metrics reflect firms' activities that ultimately contribute to value creation. The model put forward by Kaplan and Norton also implies that value creation is sequential, with some metrics being leading indicators of value creation while others are confirmatory. For instance, the quality of a firm's human capital, which contributes to learning and growth within the organization, enables the firm to improve its various internal processes (e.g., innovation), thus providing a better value proposition to customers. Higher profitability is the ultimate outcome of this chain of actions.
The value creation mapping that underlies Kaplan and Norton's (2004) model has conceptual foundations in both economics (e.g., agency theory) and management (e.g., goal theory). It has been shown analytically that value creation can be inferred from both financial measures and non-financial measures that reflect the different perspectives of managerial action (e.g., Holmstrom, 1982; Locke and Latham, 1990). There is also growing empirical validation of the Kaplan and Norton framework. For example, from a learning and growth perspective, Lin and Lin (2006) show that employee learning and training as well as teamwork are key drivers underlying customer value creation in firms.
From an internal perspective, Clarkson et al. (2004) show that better environmental performance, a key process for society, ultimately translates into value-added capital expenditures by pulp and paper firms. In terms of innovation, Xu et al. (2007) show that biotech firms with more extensive drug development portfolios have enhanced revenue opportunities and, consequently, higher stock market valuations.
From a customer perspective, recent findings suggest that customer satisfaction and loyalty are useful predictors of a firm's future financial performance and, ultimately, value creation. Smith and Wright (2004) report that product value attributes directly and differentially influence levels of customer loyalty as well as the prevailing average selling prices. Furthermore, measures of customer loyalty explain levels of relative revenue growth and profitability, and relatively high customer loyalty engenders a competitive advantage in the PC industry. In a multi-industry context, Ittner and Larcker (1998) show that customer satisfaction leads to increased revenues and, finally, to enhanced share values.
From a financial perspective, the findings of Said et al. (2003) support the contention that firms employing a combination of financial and non-financial performance measures in their compensation contracts have significantly higher mean levels of returns on assets and higher levels of market returns. Moreover, within the context of a financial institution, the results from Magnan and St-Onge (2005) indicate that the adoption of a profit-sharing plan leads to cost reductions, better asset utilization and enhanced revenue growth, which then lead to higher returns on assets.
Disclosure About Value Creation
For investors to assess if there is value creation within an organization, and fully value a firm's shares, there needs to be reliable and relevant disclosure by the organization. Thus, a critical question is why do firms choose to disclose, or not disclose, specific aspects of their value creation process.
Prior research on corporate disclosure comprises three main streams: voluntary financial disclosure, mandated financial disclosure, and non-financial disclosure such as social and environmental disclosure. Voluntary disclosure research focuses on the determinants that drive firms to release specific financial information about their underlying value to capital markets participants, typically either management earnings forecasts (Clarkson et al., 1992, 1994; Gramleich and Sorensen, 2004) or financial statement footnotes (Scott, 1994). Most of these studies adopt an information costs and benefits approach to understand firm's disclosure practices. Research on mandated disclosure focuses on how firms react to new disclosure standards or manage their disclosure in relation to mandatory requirements. For example, Balsam et al. (2003) find that the nature of the disclosure, that is, mandated or voluntary, does make a difference in how firms measure stock-based compensation expenses, an item that significantly affects earnings and, ultimately, firm value (Aboody et al., 2004). Finally, there is an extensive literature that investigates the content and drivers underlying corporate social and environmental reporting. Such disclosure is usually voluntary and non-financial in nature. In a review paper, Berthelot et al. (2003) conclude that the determinants of environmental disclosure relate to firm size, the extent of information needs by a firm's shareholders and potential litigation costs surrounding environmental obligations. However, Gray and Bebbington (2007) highlight that a relatively small proportion of firms that are listed worldwide provide corporate social responsibility (CSR) disclosure. Moreover, they argue that the average quality of disclosure about CSR activities is so uneven as to be useless for meaningful analyses and comparisons.
Web-Based Disclosure
Most prior disclosure research relies on traditional means of diffusion, that is, paper-based environmental and social responsibility reports (e.g., Neu et al., 1998) or on financial statement disclosure (e.g., Hope, 2003). However, the Web is now perceived as the best platform for the disclosure of financial and non-financial information for stewardship purposes (Lymer, 1997; Robb et al., 2001; Patten, 2002; Marston and Polei, 2004). Because it facilitates direct contact between a firm and its stakeholders, firms are able to better control their reporting strategies as they are less dependent on intermediaries such as journalists or financial analysts for the diffusion of their message (Lymer, 1999). Furthermore, information voluntarily disclosed by a firm via its website is not currently subjected to any specific regulation, unlike financial statements, proxy statements or MD&A.
Prior research allows for the identification of key determinants that underlie web-based disclosure. For instance, Ashbaugh et al. (1999) find that managers who view communications with potential shareholders as important are more likely to consider as important Internet financial reporting. Other key determinants include firm size (e.g., Debreceny et al., 2002; Ettredge et al., 2002; Patten, 2002; Marston and Polei, 2004), information asymmetry between management and investors (Ettredge et al., 2002; Cormier et al. 2009a), ownership status (Marston and Polei, 2004) and a foreign stock listing (Debreceny et al., 2002; Marston and Polei, 2004).
While the Internet is widely recognized as a flexible and versatile communications medium, most prior research on web disclosure focuses solely on financial statements or on investor-related information drawn from investor relations links on corporate websites (e.g., Ashbaugh et al., 1999; Debreceny et al., 2002; Ettredge et al., 2002; Marston and Polei, 2004; Bollen et al., 2006). With respect to web-based social disclosure, Patten (2002) shows that web-based disclosure of social responsibility information is quite low. Further, and importantly, from a social balance perspective, the web innovators in terms of product marketing are not industry leaders in terms of information disclosure. Furthermore, Cormier et al. (2009a) investigate the impact of social and human capital disclosure quality on information asymmetry. Overall, their results suggest that quantitative disclosure reduces share price volatility and increases Tobin's Q. However, despite the scope of information provided on many websites, our understanding of the strategies underlying firms' overall Internet disclosure is limited.
MODEL DEVELOPMENT
A Monitoring Perspective on Disclosure
We argue that three monitoring considerations underlie a firm's web-based disclosure about its value-creating activities. First, external parties such as investors and financial analysts incur costs to monitor if the firm is taking appropriate actions to create long-term value. Such monitoring costs mostly relate to the investigation and gathering of relevant and reliable information. Hence, if a firm releases more information, it may reduce the monitoring costs incurred by investors and other capital markets participants (Verrecchia, 1983). Most prior empirical evidence is consistent with monitoring costs driving the extent of a firm's voluntary disclosure (see a review of such literature in Healy and Palepu, 2001).
Second, concurrently to financial market monitoring, a firm's governance structure affects its disclosure, which is subject to managerial discretion. For instance, Markarian et al. (2007) show that there is convergence of disclosure and governance practices among large international firms. Hence, through various internal governance mechanisms, such as its audit committee, a firm influences the relevance and reliability of the information that it discloses (Vafeas, 2005). More specifically, firms with effective boards use disclosure as a complementary and potentially powerful governance mechanism (Zeckhauser and Pound, 1990; Craighead et al., 2004).
Finally, as primary institutional intermediaries, public media function as an additional monitoring layer which may interact with and reinforce other monitoring mechanisms. Institutional intermediaries (both expert and more general institutional intermediaries like public media) are entities that specialize in disseminating information about organizations or in evaluating their outputs (Fombrun, 1996; Rao, 2001). Intermediaries play a pivotal monitoring role to the extent that they are believed to have superior ability to access and disseminate information by virtue of their institutional roles or structural positions (Rao, 1998, 2001). The actions and choices of such intermediaries are closely followed and highly influential because of their perceived expertise in evaluating firms. In that vein, public media not only provide channels for information pooling, but also shape normative pressures through their evaluative activities. Hence, disclosure both affects and is affected by the media (Deegan and Rankin, 1996; Neu et al., 1998; Aerts et al., 2008; Aerts and Cormier, 2009).
Financial Market Monitoring
By reassuring a firm's investors regarding various aspects of its operations or performance, expanded disclosure leads to a reduction in information asymmetry between managers and investors and, ultimately, to a reduction in monitoring costs to be incurred by investors (e.g., Kim and Verrecchia, 1994). Disclosure is also beneficial to a firm as it lowers its cost of capital, raises its valuation multiples, increases its stock liquidity and enhances its interest to institutional investors (Healy et al., 1999).
However, despite investors' information needs, the decision by a firm's management to disclose information about its underlying performance is likely influenced by a trade-off between the direct costs to be incurred for providing such disclosure, the benefits to be derived by the firm or its shareholders from such disclosure and the costs of releasing private information (Scott, 1994). Hence, a firm may decide to voluntarily disclose information if it is less costly than having market participants gather the information (Milgrom, 1981; Atiase, 1985; Roberts, 1992; Lang and Lundholm, 1993).
Monitoring costs to be incurred by capital markets participants can be inferred from a firm's financial condition, systematic risk, reliance on capital markets and regulatory oversight (e.g., Scott, 1994). In turn, hypotheses are developed for each factor.
Systematic risk is a finance-derived measure of how a firm's stock price behaves compared to the stock market as a whole (otherwise called Beta). More specifically, it is the variability of a firm's stock market value relative to market-wide variability. The higher a firm's systematic risk, the more difficult it is for investors to assess precisely a firm's value. A firm's systematic risk can be a proxy for information asymmetry between the firm and investors (Leuz and Verrecchia, 2000; Botosan and Plumlee, 2005). In such a context, to reduce information asymmetry and associated monitoring costs incurred by investors and analysts, firms need to provide more disclosure. Hence, a positive relationship is expected between systematic risk and web-based disclosure:
H1a: A firm's systematic risk is positively related to its web-based disclosure.
Firms in which expansion is dependent upon continuous access to capital markets have incentives to reduce information asymmetry between managers and investors, as information asymmetry translates into higher monitoring costs for investors. A reduction in information asymmetry, through enhanced disclosure, leads to lower financing costs (Clarkson et al., 1994; Frankel et al., 1995). Lang and Lundholm (1993) document a positive relationship between capital markets reliance and voluntary disclosure. In contrast, firms with extensive free cash flows face less external monitoring (Jensen, 1986). Hence, the following hypothesis:
H1b: A firm's (lack of) reliance on capital markets is (negatively) positively related to its web-based disclosure.
Many Canadian firms have a listing on a U.S. stock market (New York Stock Exchange, Amex or NASDAQ). U.S.-based investors who follow these firms have to engage in more extensive monitoring activities as (a) information channels about these firms are foreign and more distant, and (b) the regulated levels of disclosure are lower in Canada (Leuz et al., 2003). Hence, U.S. listing is meant to capture pressures for quality disclosure (Leuz and Verrecchia, 2000). Debreceny et al. (2002) find that listing on a US exchange influences Internet financial reporting. Hence, the following hypothesis:
H1c: A firm's U.S. listing is positively related to its web-based disclosure.
By providing voluntary disclosure, firms may meet capital markets participants' information needs, and reduce their associated monitoring costs. However, firms in poor financial condition face intense monitoring by debt rating agencies and banks as well as numerous contractual constraints that limit managerial discretion (e.g., bond covenants). In such a context, firms need to ramp up their internal monitoring costs through, for example, more intensive internal audits or controls (Carcello et al., 2005). Hence, there is likely to be less external demand for information for monitoring purposes. Thus, consistent with prior findings (McGuire et al., 1988; Cormier and Magnan, 2003), it is expected that there is a negative relationship between a firm's financial condition, as proxies by its leverage, and performance disclosure.
H1d: A firm's leverage is negatively related to its web-based performance disclosure.
Governance Monitoring
The quality of a firm's governance and the extent of its voluntary corporate disclosure are closely intertwined (Bushman and Smith, 2001). Governance encompasses monitoring and incentive practices that ensure managerial actions are consistent with shareholder interests. Within a governance framework, three major monitoring mechanisms influence corporate disclosure decisions: owners, directors and auditors. In addition, the overall quality of a firm's governance can also be assessed by observing executive compensation, one of the few visible outcomes of board governance. For instance, Bebchuk (2004) labels executive compensation the ‘smoking gun’ of corporate governance.
Ownership structure can determine the level of monitoring and, thereby, the level of disclosure (Zeckhauser and Pound, 1990; Eng and Mak, 2003). Usually, the need for external monitoring is reduced in firms with concentrated ownership. More specifically, since the dominant shareholders have access to the information they need, closely held firms are expected to be unresponsive to public investors' monitoring costs (Hope, 2003) or other stakeholders' needs (Roe, 2003). Hence, the following hypothesis:
H2a: Concentrated ownership is negatively related to web-based disclosure.
Non-executive directors on a board enhance the monitoring of corporate insiders (Fama and Jensen, 1983). For instance, Chen and Jaggy (2000) document a positive relationship between board composition (proportion of independent directors) and the comprehensiveness of information in mandatory financial disclosures by Hong Kong firms. Karamanou and Vafeas (2005) provide evidence that firms with higher quality governance characteristics are likelier to issue voluntary earnings forecasts.
In addition to board composition, Chtourou et al. (2004) find that board size is associated with less earnings management, that is, higher quality disclosure. Finally, boards are more likely to be effective in overseeing management if they meet regularly (Vafeas, 2005). Hence, board monitoring effectiveness relies on board composition (i.e., outside vs inside directors), board size and level of activity.
H2b: Board effectiveness is positively related to web-based disclosure.
In Canada, audit committees are mandatory and take over a significant portion of the board's responsibility with respect to disclosure. By regulation, audit committees must comprise at least three members who are independent from management. We argue that three is a small number for the audit committee to play effectively its monitoring role and that adding a few more members could be beneficial in that regard. Effective monitoring also implies a minimum number of meetings by the audit committee (Abbott et al., 2003). With respect to disclosure, Bronson et al. (2006) find that voluntary management reports on internal controls are more likely for firms that have an audit committee that meets more often. Hence, the following hypothesis:
H2c: Audit committee effectiveness is positively related to web-based disclosure.
Finally, there is evidence that compensation such as stock options can align manager interests with shareholder interests (Hanlon et al., 2003). However, contracting costs may lead to incomplete contracts and agency conflicts. Aboody and Karnak (2000) show that managers may mislead shareholders by accelerating bad news and by delaying good news to reduce the exercise price of stock option grants. Hence, CEOs with extensive stock option values are likely to be opportunistic in their disclosure strategies. Therefore, no directional prediction is made.
H2d: A CEO's stock option holdings relate to a firm's web-based disclosure.
Media Monitoring
The impact of public media on corporate disclosure derives primarily from their ability to focus public attention on specific firms or issues (Deephouse, 2000; Pollock and Rindova, 2003; Aerts et al., 2008). By setting the public agenda, the media do not necessarily mirror public concerns (Ader, 1995). However, through the transfer of salience, the media agenda may transpose into the public agenda (Caroll and McCombs, 2003). Therefore, the media are actively involved in the construction of social impression processes and related monitoring (Gamson et al., 1992). We expect both a direct and an indirect effect of media coverage on disclosure. The direct effect results from media coverage voicing social concerns driving the corporate transparency agenda, while the indirect effect stems from its interaction with governance monitoring. Thus, the following hypothesis:
H3a: The extent of media coverage is positively related to web-based disclosure.
However, we expect that in the absence of media coverage, the impact of external board members on disclosure transparency will be restricted to financial disclosure while they will influence non-financial disclosure when the firm is facing extensive media coverage. Public rhetoric surrounding the role and functioning of independent directors may have created an ‘expectations gap’ (Reay, 1994; Hooghiemstra and Van Menem, 2004), that is, the existence of a gap between what independent directors can reasonably be expected to accomplish and what is expected by the external business environment.1 This expectations gap puts pressure on independent directors if these expectations are effectively ‘voiced’. One could refer to these pressures as external accountability pressures. External (independent) members have their reputation as professional referees at stake, and will be more sensitive to reputational risk threats than inside directors (Aguilera, 2005).
We expect that media monitoring, as measured by media exposure, has a positive impact on the role played by independent directors concerning a firm's disclosure.
H3b: The extent of media coverage enhances the effect of an independent board of directors on a firm's web-based performance disclosure.
METHOD
Sample
The sample comprises 155 Canadian publicly traded firms. All non-financial firms represented on the Toronto Stock Exchange S&P/TSX 300 Index were initially identified in 2002.2 For the current research, web-based disclosure was collected from websites (web page and HTML) in summer 2003 and 2005. Multivariate analyses were performed on 2005 web disclosure. Financial data for 2004 was collected from the Stock Guide and governance data was collected from 2004 proxy statements. The final sample is 139 firms since, out of the initial sample of 155 firms, there are missing data for board size and independence (9 firms), stock options (5 firms), and debt and stock issues (2 firms).3 Sample firms operate in the these industries: metals and mines; gold and precious metals; oil and gas; paper and forest products; consumer products; industrial products; real estate; utilities; communication and media; merchandising.
Empirical Approach
We explore the determinants of web-based disclosure by using the following model:
New disclosure and governance requirements came in force in March 2004 (Canadian Securities Administrators, 2004). Since some information related to corporate governance was not available before 2004, we restrict our regression analyses to the determinants of web-based disclosure in 2005.
Measurement of Web-Based Disclosure
We focus on voluntary disclosure available from a firm's website in HTML format since it is comprehensive and accessible to all shareholders at a low cost. In a Canadian context, however, mandated disclosure is filed on SEDAR (a system of electronic data archiving and retrieval that is maintained by securities regulators). The SEDAR website (http://www.sedar.com) comprises all documents in which disclosure is mandated by securities regulators: financial statements, annual reports, proxy statements, MD&A and press releases. However, all these documents are also available on paper and most investors still receive them in paper form. The content of these documents is regulated. Therefore, our disclosure measure reflects only information that the firm has voluntarily decided to post on the web.
Disclosure indicators, financial or non-financial, are based on balance scorecard literature and emerging performance measurement practices (e.g., Standard & Poors, 2002, for financial and governance disclosure; Pirchegger and Wagenhofer, 1999; Patten, 2002; and Marston and Polei, 2004, for investors, governance and social responsibility disclosures; Kaplan and Norton, 1996; Ittner and Larcker, 1998; and Robb et al., 2001, for indicators about operations' efficiency, value for client, innovation, development and growth). We analyse web-based disclosure by coding the content of corporate websites using a grid developed by Cormier et al. (2009b) (see Appendix).4 The grid comprises 111 items, which are grouped in eight categories: website quality, financial performance, corporate governance, customer value, human and intellectual capital, production efficiency, innovation, development and growth, and social responsibility. Website quality is captured by features such as interactive components, video-audio access, hyperlinks, etc. These features may enhance firms' ability to convey their message to outside users. Our measure of website quality is consistent with Marston and Polei (2004) who, in their assessment of web disclosure strategy, consider the timeliness of the information, technological features (e.g., loading time) and navigation support (e.g., convenience and usability). The rating is based on a score of 1 to 3 per element, with each item possibly comprising many elements. A score of 3 is given for an element that is described in quantitative terms (‘hard’ information), a score of 2 when an element is described specifically and a score of 1 for an element discussed in general (‘soft’ information). For instance, within the customer profile item, we may find many elements of information on market segments, market shares, number of customers, etc. Each of these elements could be coded, which would imply a total score for this item greater than 3. We decided not to cap individual scores since our grid aims to capture all relevant information excluding any overlap or repetitions. Such a coding scheme, with higher weights for hard disclosure items than for soft disclosure, is consistent with prior research (Cho and Patten, 2007; Clarkson et al., 2008). The use of a coding scale to qualify a firm's disclosure is consistent with prior work in financial/non-financial disclosure (e.g., Gibbins et al., 1990; Lang and Lundholm, 1993; Botosan, 1997; Healy et al., 1999; AIMR, 2002; Robins and Stylianou, 2003).
The coding was performed concurrently by two graduate students. Coding instructions, as well as standardized coding worksheets, were prepared beforehand. Each coder then applied the following coding sequence: (a) independent identification of the occurrence of items relative to the different coding categories; (b) independent coding of the items according to quality level of content; and (c) timed reconciliation on a subset of company reports. The coders were trained by one of the co-researchers in applying coding instructions and in using the coding worksheets. The training went over a few weeks as the coders and the researcher mutually verified their coding on a subsample of test cases, typically two firms by industry. Coders were unaware of the research hypotheses. Initial differences in identifying grid items accounted for an average of 7% of the maximum number of items identified. Of the information quality level coding, less than 10% had to be discussed for reconciliation. Disagreement between coders mostly happened at the beginning of the coding process (essentially the first forty sample firms). The researcher reconciled coding disagreements exceeding 5% of the highest total score between the two coders. Smaller disagreements were resolved by the two coders themselves. Internal consistency estimates (Cronbach's alpha on score components) show that the variance is quite systematic (alpha= 0.713 in 2005, see Table 2). This is slightly higher than Botosan (1997), who finds an alpha of 0.64 for an index including five categories of disclosure in annual reports.5
| 2003 | 2005 | |||
|---|---|---|---|---|
| Mean score | Cronbach's alpha | Mean score | Cronbach's alpha | |
| Web-quality | 14.80 | 0.649 | 17.47 | 0.661 |
| Financial performance | 8.03 | 0.865 | 6.18 | 0.869 |
| Corporate governance | 17.14 | 0.709 | 17.93 | 0.830 |
| Customer value | 17.35 | 0.794 | 14.86 | 0.703 |
| Human / intellectual capital | 9.27 | 0.823 | 9.88 | 0.761 |
| Production efficiency | 9.44 | 0.623 | 7.32 | 0.648 |
| Innovation / development and growth | 3.15 | 0.819 | 2.41 | 0.735 |
| Social responsibility | 12.59 | 0.765 | 16.05 | 0.704 |
| Total score | 91.84 | 0.644 | 92.17 | 0.713 |
| N | 167 | 155 | ||
For analysis purposes, a firm's disclosure score is scaled by its industry median disclosure score. Two reasons motivate that empirical choice. First, there are industry-specific disclosure patterns (e.g., Aerts et al., 2006). Second, adding industry-specific dummy variables would use up many degrees of freedom and seriously compromise our analyses' statistical power. Hence, it is deemed appropriate to control for industry patterns by scaling the dependent variable by industry mean.
Measurement of Web-Based Disclosure Determinants
Financial market monitoring Five variables are used to capture capital markets' information considerations that affect disclosure: Systematic risk; New financing; Free cash flow; U.S. listing; and Leverage. Systematic risk is measured by Beta (extracted from Stock Guide). A positive relationship is expected between systematic risk and the extent of disclosure. New financing is a measure of external financing or reliance on capital markets. It can be proxied by issues of long-term debt and equity (Dechow et al., 1996). The variable measures the actual amount of long-term financing raised through stock or debt offerings scaled by total assets. Free cash flow is a second measure of external financing (Jensen, 1986; Dechow et al., 1996). It proxies for the demand for external financing (in negative sense) by measuring a firm's ability to cover its capital expenditures. The higher the free cash flow, the lower the need for external financing. We measure that variable as cash flow from operations in 2004 minus the average of capital expenditures from 2002 to 2004 scaled by total assets. U.S. listed firms are meant to face disclosure pressures internationally (Leuz and Verrecchia, 2000). Debreceny et al. (2002) find that in addition to a firm's size, listing on U.S. exchange is a specific determinant of Internet financial reporting. Hence, U.S. listing is introduced as a binary variable (1; 0 if not) and a positive relation is expected between SEC and disclosure. Leverage is measured by long-term financial debt on stockholders' equity.
Governance monitoring comprises five variables: Concentrated ownership; Board independence; Board size; Audit committee size; and CEO stock options. Concentrated ownership is measured as a dichotomous variable taking a value of one (1) when an investor, or a related group of investors, owns more than 10% of a firm's outstanding voting shares, and zero (0) otherwise. Ten per cent is the ownership level that must be disclosed in Canada. A negative relationship is expected between concentrated ownership and disclosure. Board effectiveness is assumed to be captured by two distinct variables, Board independence and Board size. Board independence is a summary variable that reflects two facets of a board's composition. A board is deemed to be independent (dependent) if the proportion of outside directors, that is, directors who are not executives and/or their relatives and controlling stockholders, exceeds 50% (is less than 50%). Another aspect of board independence is the separation of the roles of chair and chief executive officer. The variable takes the value of zero (0) when the majority of directors are not independent, a value of one (1) when the majority of directors are independent, and a value of two (2) when the majority of directors are independent and the functions of CEO and chair of the board are separate. We expect a positive relationship between this variable and disclosure. Board size is measured by the number of its members. We expect a positive relationship between this variable and disclosure. In modern governance, monitoring is increasingly performed by board committees. With respect to disclosure, the audit committee is the pivotal governance mechanism. Hence, to complement board effectiveness, we consider also audit committee effectiveness, which is measured by Audit committee size, that is, its number of members. Finally, CEO stock options reflect the intrinsic value of unexercised stock options held the CEO as per the proxy statement date, divided by his/her CEO cash compensation.
Media monitoring Media monitoring upon a firm is proxied by Media exposure. It is computed by taking the average number of articles for the period 2000 through 2004, as contained in the ABI Disclosure database. The reason for this choice is that disclosure in 2005 may be affected by the amount and types of articles published about a firm in the recent past. We expect that as media exposure increases, the firm will increase its disclosure.
Control variables Three control variables are added to the empirical model: Firm size; High skill employees; and Repeat customer relations. Prior evidence is consistent in highlighting a positive relation between the extent of corporate disclosure and Firm size, which is measured as the Ln (Total Assets) (e.g., Scott, 1994)). An indicator variable captures employees' skill level (High skill employees = 1, 0 otherwise dependent upon the type of labour required by the industry in which the firm operates). Employees' perceived skill levels enhance their potential visibility and the firm's dependence upon them, thus providing the firm with an incentive to be more transparent in its disclosure. For instance, several media recognize and publish rankings of ‘Employers of choice’. Typically, many of these firms will be high value-added firms with large professional or scientific labour forces.6
We also consider that a firm's relations with clients may affect their disclosure, with firms that engage in long-term relationships with their customers being more likely to provide more web-based disclosure than firms where such long-term commitments do not exist (e.g., Bowen et al., 1995). Sample firms are classified into two groups, with the first group comprising firms with short-term repeated transactions with clients (e.g., grocery stores) while the second group of firms maintain long-term relationships with its clients (e.g., cable and entertainment) or engage in long-term contracts with them (e.g., durable goods with warranties). The variable Repeat customer relations is coded (1) for short-term or repeat relations with customers, or (0) otherwise.
RESULTS
Descriptive Statistics
Table 1 provides some descriptive statistics about explanatory variables as well as cross-correlations. Overall, sample firms are relatively large (average assets of $4.3 billion) and exposed to media (average of 4.7 articles per year over the last 5 years). Fifty-nine per cent of sample firms have a concentrated ownership, while 45% are listed in the United States. CEO stock option value in-the-money represents 1.79 times their salary and bonus. All correlations are below the 0.50 threshold, except for three variables. The highest correlations are between Systematic risk and firms with High skill employees (0.55), Board size and Audit committee size (0.55), and Board size and Firm size (0.53). This suggests that collinearity is not a concern.
| Mean | STD | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Leverage | 0.21 | 0.16 | 1 | *−0.12 | *0.17 | 0.09 | −0.02 | 0.07 | *−0.14 | *0.21 | *0.15 | −0.11 | −0.06 | *0.21 | *−0.27 | −0.02 |
| 2 | Systematic risk | 0.68 | 0.49 | 1 | −0.01 | −0.01 | 0.08 | −0.05 | 0.06 | *−0.12 | *−0.13 | *0.21 | *0.29 | *−0.12 | *0.55 | *0.11 | |
| 3 | New financing | 0.09 | 0.12 | 1 | −0.09 | *0.23 | −0.06 | *−0.13 | 0.01 | 0.09 | −0.04 | 0.02 | −0.06 | −0.01 | 0.02 | ||
| 4 | Free cash flow | 0.02 | 0.15 | 1 | 0.01 | −0.02 | *−0.12 | *0.22 | *0.19 | 0.05 | 0.03 | *0.30 | −0.04 | 0.07 | |||
| 5 | U.S. listing | 0.45 | 0.49 | 1 | *−0.20 | 0.01 | 0.01 | 0.02 | *0.15 | *0.12 | 0.01 | 0.07 | *−0.19 | ||||
| 6 | Concentrated ownership | 0.59 | 0.49 | 1 | −0.10 | 0.07 | −0.08 | −0.01 | *−0.14 | −0.04 | −0.08 | *0.12 | |||||
| 7 | Board independence | 0.91 | 0.51 | 1 | 0.11 | 0.09 | −0.02 | −0.09 | −0.07 | *0.18 | *−0.16 | ||||||
| 8 | Board size | 9.99 | 2.76 | 1 | *0.55 | −0.07 | *0.19 | *0.53 | *−0.15 | −0.08 | |||||||
| 9 | Audit committee size | 3.98 | 1.10 | 1 | 0.01 | *0.16 | *0.42 | *−0.19 | 0.06 | ||||||||
| 10 | CEO stock options | 1.79 | 21.70 | 1 | 0.10 | 0.11 | 0.10 | −0.02 | |||||||||
| 11 | Media exposure | 4.70 | 9.71 | 1 | *0.33 | *0.14 | 0.01 | ||||||||||
| 12 | Size ($million) | 4,268 | 7,016 | 1 | *−0.24 | *−0.16 | |||||||||||
| 13 | High skill employees | 0.18 | 0.38 | 1 | *0.20 | ||||||||||||
| 14 | Repeat customer relations | 8 873 | 16 071 | 1 |
- * Significant at 0.10 two-tailed
As illustrated in Table 2, total web-based disclosure does not vary significantly from 2003 (mean score of 91.84) to 2005 (mean score of 92.17). Among the eight disclosure components, we observe increases in web quality and in social responsibility disclosure and the opposite concerning customer value, production efficiency and innovation, development and growth.
Multivariate Results
Determinants of web-based disclosure Table 3 provides evidence regarding the determination of web-based disclosure.7 The Total score column shows results from a cross-sectional OLS regression between overall disclosure (dependent variable) and variables proxying external for monitoring costs, governance and media monitoring, as well control variables. The following eight columns provide results for different facets of disclosure. Focusing on the regression for total disclosure, its explanatory power is 45.3% (p < 0.000). Diagnostic procedures (VIF and normality tests) do not reveal multicollinearity or normality problems (variance of inflation factors do not exceed three in any regression). The use of the Belsh-Kuhley procedure allows for the identification of three outliers (absolute value of standardized DFITS > 1.0).8 Hierarchical regressions indicate that both control and experimental variables contribute in explaining disclosure. Adding experimental variables to control variables generates a significant F change for total disclosure model(3.120; p < 0.001). F change tests are significant for seven out eight disclosure components.
| Explanatory variables | # | Total score | Web-quality | Financial performance | Corporate governance | Customer value | Human, intellectual capital | Production efficiency | Innovation/development and growth | Social responsibility |
|---|---|---|---|---|---|---|---|---|---|---|
| Intercept | −2.063*** | 0.306 | 3.663 | **−0.385 | 0.224 | ***−6.348 | −1.629 | 2.945 | ***−9.357 | |
| Financial market monitoring | ||||||||||
| Systematic risk | + | **0.117 | −0.006 | −0.255 | 0.069 | **0.198 | **0.372 | ***0.654 | ***2.560 | 0.087 |
| New financing | + | −0.209 | 0.003 | 0.590 | 0.230 | **−0.772 | −0.552 | −0.953 | 0.223 | −1.001 |
| Free cash flow | − | *−0.031 | *−0.081 | 1.204 | *−0.117 | 0.204 | *−0.353 | −0.151 | 0.642 | 0.209 |
| U.S. listing | + | ***0.294 | 0.011 | *0.510 | ***0.363 | ***0.211 | ***0.327 | *0.256 | *0.796 | ***1.124 |
| Leverage | − | ***−0.534 | −0.053 | **−1.985 | −0.181 | −0.158 | ***−1.302 | −0.469 | *−2.108 | **−2.683 |
| Governance monitoring | ||||||||||
| Concentrated ownership | − | *−0.055 | ***−0.062 | *−0.243 | ***−0.247 | **0.174 | −0.001 | 0.102 | ***−1.879 | 0.325 |
| Board independence | + | **0.100 | −0.020 | **0.650 | **0.209 | −0.105 | 0.027 | −0.189 | 0.129 | ***0.652 |
| Board size | + | ***0.036 | ***0.011 | ***0.172 | *0.031 | −0.010 | ***0.074 | *0.049 | ***0.347 | 0.025 |
| Audit committee size | + | 0.013 | −0.004 | ***−0.467 | −0.011 | *0.064 | 0.047 | 0.058 | *−0.407 | 0.209 |
| CEO stock options | ? | ***−0.003 | −0.001 | −0.003 | −0.002 | −0.001 | 0.001 | ***−0.010 | **0.019 | **−0.012 |
| Media monitoring | ||||||||||
| Media exposure | + | −0.001 | **−0.003 | −0.025 | **−0.016 | **−0.011 | *0.009 | −0.001 | ***−0.114 | *0.027 |
| Control variables | ||||||||||
| Firm size | + | ***0.119 | ***0.029 | −0.090 | ***0.187 | 0.023 | ***0.293 | 0.091 | −0.148 | ***0.430 |
| High skill employees | + | 0.019 | 0.028 | −0.471 | −0.013 | *0.126 | *0.308 | 0.174 | 0.253 | −0.485 |
| Repeat customer relations | + | 0.050 | *0.025 | 0.306 | **0.185 | 0.023 | 0.006 | −0.245 | 0.329 | 0.272 |
| Adjusted R2 | 45.3% | 12.2% | 6.0% | 19.5% | 13.5% | 36.6% | 11.6% | 27.8% | 29.5% | |
| F statistic | 9.12 | 2.34 | 1.62 | 3.34 | 2.42 | 6.48 | 2.29 | 4.66 | 5.03 | |
| P value | (0.000) | (0.007) | (0.085) | (0.000) | (0.005) | (0.000) | (0.009) | (0.000) | (0.000) | |
| Instrumental variables | 3.120 | 1.789 | 2.075 | 2.246 | 1.582 | 2.259 | 1.871 | 3.115 | 2.524 | |
| F-change | (0.001) | 0.072) | (0.032) | (0.020) | (0.122) | (0.019) | (0.055) | (0.001) | (0.009) | |
| N= 139 |
- * p < 0.10, ** p < 0.05, *** p < 0.01. One-tailed if there is a predicted sign and in the right direction, two-tailed otherwise.
- # Predicted sign of the relation between an explanatory variable and the dependent variable.
- Dependent variable: Disclosure score over industry median.
Among variables proxying for external monitoring costs, results show that, to reduce information asymmetry and associated monitoring costs incurred by investors, firms provide more disclosure when facing high environmental uncertainty as expressed by Systematic risk (H1a) (0.117; p < 0.050). In contrast, firms with extensive free cash flows face less external monitoring and hence disclose less information than others (H1b) (−0.031; p < 0.100). As expected, it appears that high leverage reduces external demand of information for monitoring purposes with (H1d) (−0.534; p < 0.010). Furthermore, results show that governance and monitoring issues are associated with disclosure. More specifically, Concentrated ownership translates into less disclosure (H2a) (−0.055; p < 0.100). In addition, board effectiveness, as measured by Board independence (0.100; p < 0.050) and Board size (0.036; p < 0.010), is positively related to disclosure (H2b). Finally, the magnitude of the CEO's stock option holdings is negatively associated with disclosure (H2d) (−0.003; p < 0.010). These results suggest that efficient governance leads to more transparency while the extent of CEO stock options leads to less transparency. Despite our expectations (H3), there is no significant relationship between media exposure and web-based disclosure. Among control variables, we find a positive relationship between Firm size (0.119; p < 0.010) and disclosure. Moreover, firms listed on U.S. stock exchange (SEC) (0.294; p < 0.010) provide more extensive disclosure than others do.9 Overall, there is support for the impact of governance and monitoring variables on voluntary disclosure. To get better insights into the web-based disclosure process, we now break down total disclosure into its eight components. Table 3's other columns provide evidence regarding the determination of different disclosure components, with the best explanatory powers being for disclosure about human and intellectual capital (adjusted R2= 36.6%), social responsibility (29.5%), innovation, development and growth disclosure (27.8%), and corporate governance (19.5%).
Some patterns that underlie disclosure relationships can be highlighted. First, financial market monitoring seems relevant to explain disclosure about human and intellectual capital (H1a, H1b, H1c, H1d), innovation, development and growth (H1a, H1c, H1d) and customer value (H1a, H1b, H1c). All these disclosures relate to intangible assets, which are often not well reflected in traditional or regulated disclosures such as audited financial statements (e.g., Lev and Zarowin, 1999). Second, governance variables appear to have particular relevance in explaining disclosure about financial performance (H2a, H2b, H2c), corporate governance (H2a, H2b) and innovation, development and growth (H2a, H2b, H2c, H2d). Third, the relationships between web-based disclosure components and information costs and benefits and governance variables are consistent with those reported for total disclosure. However, for media monitoring, two patterns emerge. On the one hand, consistent with H3a, media exposure translates into more human and intellectual capital and social responsibility disclosures. On the other hand, media exposure seems to translate into less web quality and disclosure on governance, customer value, and innovation, development and growth. This would explain the lack of a significant relation between media exposure and total disclosure, which we now explore.
Additional Governance-Focused Analyses
Reputational risk rests highly on media coverage. To assess the impact of media monitoring on the role played by external members of the board, we add an interaction term Board independence*Media exposure to the regression. Consistent with H3b, results (not tabulated) show that the coefficient for the interaction term is positive and significant (0.016; p < 0.052), while the coefficient for Board independence is not significant (−0.021; p < 0.705). The coefficient for Media exposure is negative and significant (−0.013; p < 0.024). This result suggests that media monitoring, as measured by media exposure, has a positive impact on the role played by external directors regarding disclosure transparency.
Furthermore, we split the variable Board independence in two different variables: independent members (1/0) and CEO not chair of the board (1/0). Our results (untabulated) suggest that an external board has a positive impact on disclosure only when the CEO is not the chair of the board and only in the presence of public media coverage. Hence, the coefficient for the variable CEO not Chair*Media exposure is positive and significant (0.022; p < 0.010). Coefficient for CEO not chair (0.054; p < 0.493), Independent members (−0.036; p < 0.749), and Independent members*Media exposure (0.018; p < 0.749) are not significant. This finding suggests board member independence is effective when the chair is distinct from the CEO, thus supporting Hermalin and Weisbach's (1998) argument.
One way to measure the effectiveness of a board committee is to look at the frequency of its meetings. As an additional sensitivity analysis, we add board of director meetings as well as audit committee meetings to our model. Concerning the audit committee, best practices suggest three or four meetings per year (KPMG, 1999). We expect that the disclosure of the frequency of board meetings and audit committee meetings will be positively related to disclosure. We observe that twenty-three out of our sample firms did not disclose this information. Results for the reduced sample (not tabulated) show a positive association between board meetings and financial performance disclosure (0.096; p < 0.010) and a negative association with disclosure about operations efficiency (−0.041; p < 0.050). In addition, more audit committee meetings translate into more total disclosure (0.018; p < 0.100), as well as disclosure on innovation, development and growth (0.213; p < 0.050) and governance (0.040; p < 0.100).
Analyses Controlling for Endogeneity
Prior research documents that the proportion of independent directors can be endogenously determined. Peasnell et al. (2005) find that the proportion of independent board members is associated with block ownership, firm size and leverage. Since a firm's communication strategy may affect disclosure, media exposure and board independence simultaneously, we first assess whether or not interrelations exist between these variables using the Hausman test. Using this procedure, we reject the null hypothesis of no endogeneity with respect to disclosure and board independence (t=−2.630; p < 0.010), and between media exposure and board independence (t= 5.119; p < 0.000). To assess whether our results could be biased by the presence of endogeneity, we adopt the following simultaneous equations model using a three-stage least square approach:
The board independence regression model presented in Table 4 has an explanatory power of 17.8% (F statistic p value < 0.000). The coefficients are statistically significant, that is, Free cash flow (−0.524; p < 0.100 two-tailed) and Lag disclosure (0.655; p < 0.100 two-tailed).10 The media exposure regression model has an explanatory power of 16.7% (F statistic p value < 0.000). Three out of four coefficients are significant, that is, Number of employees (0.001; p < 0.010), SEC (4.335; p < 0.010) and High skill employees (4.784; p < 0.010). Concerning the disclosure regression model, the explanatory power is 32.1% (F statistic p value < 0.000). Results remain similar to those estimated with an OLS regression (Table 3) except for two variables. The coefficients for Systematic risk and Concentrated ownership become insignificant. These findings do not suggest endogeneity problems.
| Disclosure | Board independence | Media exposure | |
|---|---|---|---|
| Systematic risk | 0.076 | 0.055 | |
| Free cash flow | 0.397 | *−0.524 | |
| U.S. listing | ***0.309 | ***4.335 | |
| Leverage | *−0.400 | −0.248 | |
| Concentrated ownership | −0.014 | −0.006 | |
| Board independence | *0.889 | ||
| Board size | *0.025 | ||
| Audit committee size | −0.018 | ||
| CEO stock options | *−0.002 | ||
| Media exposure | −0.024 | ||
| Firm size | ***0.140 | −0.035 | |
| High skill employees | 0.046 | ***4.784 | |
| Lag disclosure score scaled by the industry median | *0.655 | ||
| Repeat customer relations | *0.155 | −0.202 | |
| Number of employees | ***0.001 | ||
| R 2 | 32.1% | 16.7% | 17.8% |
| χ2 (p value)N= 139 | 75.1(0.000) | 26.6(0.000) | 12.3(0.000) |
- * p < 0.10, ** p < 0.05, *** p < 0.01. Two-tailed significance.
CONCLUDING REMARKS
Our paper focuses on the web disclosure being provided by Canada's leading publicly listed firms about their value creation processes. We adopt a conceptual framework that weaves together three complementary monitoring perspectives, with variables proxying for financial market monitoring, governance monitoring and media monitoring being found to underlie web-based disclosure. Results are consistent with monitoring costs faced by investors, the effectiveness of board monitoring and media monitoring driving web-based disclosure about value creation. Ourresults also suggest that disclosure complements media and board governance as a monitoring mechanism.
Our results are subject to some limitations. First, the paper focuses on web-based performance disclosure only, which represents but one facet of a firm's disclosure strategy. However, most prior research emphasizes specific dimensions (e.g., annual report, press releases, etc.) as it is difficult to construct a comprehensive measure of corporate disclosure. A second potential limitation is the paper's focus on HTML disclosure, which excludes hyperlinked documents in PDF. However, these documents (e.g., quarterly or annual financial statements, press releases, annual reports, sustainability reports or proxy statements) are typically also published in paper form (e.g., Aerts et al., 2007). Moreover, our experience and further investigations of specific firms suggest that most PDF disclosures are mandated disclosures. Another potential limitation is the use of a coding grid to quantify corporate disclosure, a measure that may be perceived as subjective. However, our coding approach is consistent with recent studies. Finally, the paper adopts a monitoring perspective to explain corporate disclosure. Alternative explanations may be provided to explain the same relations (e.g., for free cash flow). However, monitoring considerations offer a comprehensive view.
Our study has the following implications. From a theoretical perspective, we conclude that a firm's disclosure about various aspects of value creation is not solely driven by external monitoring considerations but is also influenced by the effectiveness of its internal governance mechanisms. One can infer that governance may need to be reinforced by external monitoring through more open disclosure. This finding is consistent with Craighead et al. (2004), who show that disclosure can be viewed as a governance mechanism that complements boards and owners. In addition, directors' monitoring role seems to be influenced by media exposure. This suggests that directors perceive themselves to be accountable to a wider audience than just investors. From a practical perspective, our results suggest that regulators may need to intervene in corporate disclosure practices, especially in contexts where governance is weak or less effective.
An objective for further research should be to implement a strategic watch to further investigate the importance of temporal trends and industry membership in web disclosure, as media monitoring can evolve over time and cause shifts in firms' disclosure strategies. In addition, the use and impact of web-based disclosure as a monitoring mechanism could be further investigated. For instance, what is the reliance of financial analysts on information conveyed through the Web? More in-depth analyses of patterns and determinants underlying disclosure components could also be performed. Finally, the relative adoption and use of the Web as a reporting platform could be compared across various countries, with different socio-political environments.
Appendices
AppendixPERFORMANCE DISCLOSURE GRID
| Up-to-date (less than 1 month: 3; 1 to 2 months: 2; 3 months: 1) | Increase in sales / market shares |
| Internal links (levels of internal links) | Increase in investments |
| External links (number) | Total growth |
| Refers to documents for additional information (0 or 1) | Total innovation, development et growth |
| Information usually on the web versus paper (0 or 1) | Product description |
| User friendly (high: 3; average: 2; low: 1) | Quality / up-to-date technology |
| Interactive components (0 or 3) | Reliability: errors / returns |
| Video-audio access (0 or 3) | Price |
| Total web-quality | Delivery time |
| Liquidity | Awards |
| Indebtedness | Total product |
| Interest coverage | Customer profile / market segment / market share / number of customers |
| Total solvency | Pre-sales support: information / counsel / orders follow-up |
| Net operating income | After-sales service / insurance |
| Gross margin | Customer satisfaction / complaints management |
| REA or REO | Customer loyalty |
| EPS (diluted) 1 | Awards |
| Stock price or stock return | Total customers |
| EVA | Service Internet (1 of order, 2 if service, 3 if both) |
| Total profitability | E-business sales |
| Total financial performance | E-business productivity [Cost efficiency / speed] |
| Leadership | Impact (award, number of users or visitors) |
| Mission | Total e-business |
| Strategic planning | Total customer value |
| Risk management | Hiring / new employees |
| Globalization | Qualification / expertise |
| Total strategic management | Training |
| Competence of managers | Description of job requirements 1, 2, 3 |
| Managers’ compensation | Total competence |
| Total managers | Employee empowerment / involvement |
| Competence Board | Capacity to suggest and to implement changes |
| Independence Board | Teamwork |
| Compensation (stocks/options) | Performance assessment |
| Other committees | Performance based compensation |
| Total directors | Earnings-based compensation |
| Competence Audit committee | Carrier opportunities |
| Independence Audit committee | Award |
| Relations with external auditors | Fringe benefits |
| Relations with internal auditors | Total motivation/work climate |
| Total Audit committees | Employees satisfaction, survey |
| Ownership structure | Employee turnover |
| Other | Other |
| Total ownership | Total satisfaction |
| Total corporate governance | Total human/intellectual capital |
| Investment ($) | Purchases of goods and services |
| Reengineering / downsizing | Employment opportunities |
| Process improvement methods (ex. Kaisen) | Job creation |
| ISO 9000, total quality management—TQM | Equity programs |
| Others (benchmarking, JIT, etc.) | Human capital development |
| Total operations rationalization | Regional development |
| Production cost | Gifts and sponsorships |
| Production capacity | Accidents at work |
| Waste | Health and safety programs |
| Inventory / run out rate | Product-related-incidents |
| Quality of equipment and technology | Products in development and environment |
| Flexibility | Product safety |
| Process description (1,2,3) | Business ethics |
| Others | Strategic alliances |
| Total productivity-cost | Community involvement |
| Production time | Social activities |
| Unplanned downtime | Total social responsibility |
| Total productivity-speed / cycle time | Total performance disclosure |
| Partnerships | |
| Acquisitions | |
| Total strategic alliances | |
| Total production efficiency | |
| Sales—new products | |
| Market share—new products | |
| Awards | |
| Total new products | |
| Investments in R&D | |
| Description of products in development | |
| Product testing | |
| Awards | |
| Others—R&D | |
| Total R&D |
