attributes of social and human capital disclosure and information asymmetry between managers and...

Can J Adm SciCopyright © 2009 ASAC. Published by John Wiley & Sons, Ltd. 71 26(1), 71–88 (2009)

Attributes of Social and Human Capital Disclosure and Information Asymmetry between Managers and InvestorsDenis Cormier*UQÀM

Walter AertsUniversity of Antwerp

Marie-Josée LedouxUQÀM

Michel MagnanConcordia University

Canadian Journal of Administrative SciencesRevue canadienne des sciences de l’administration26: 71–88 (2009)Published online in Wiley Interscience (www.interscience.wiley.com). DOI: 10.1002/CJAS.89

The authors acknowledge the fi nancial support of l’Autorité des marchés fi nanciers (Québec) and the Social Sciences and Humanities Research Council of Canada.*Please address correspondence to: Denis Cormier: ESG UQÀM, P.O. Box 8888, down town station, Montréal, Québec, Canada H3C 2P8. Email: [email protected]

AbstractWe extend the literature on voluntary disclosure by investigating the impact of precision attribute of social and human capital disclosure on information asymmetry. We provide evidence on how the stock market reacts to different levels of information precision. Overall, results suggest that quantitative disclosure reduces share price volatility and increases Tobin’s Q. As expected, fi rm size attenuates the impact of precision attribute of disclosure on information asymmetry. Furthermore, it appears that fi rms take into account ultimate costs and benefi ts to shareholders when determining precision attribute of their disclosure. Finally, our results suggest that effi cient governance leads to more disclosure. Copyright © 2009 ASAC. Published by John Wiley & Sons, Ltd.

JEL classifi cations: M40, M14, 015, M50

Keywords: information asymmetry, corporate governance, human capital disclosure, social capital disclosure

RésuméDans cette étude, nous contribuons aux recherches existantes sur les choix discrétionnaires des entreprises en matière de divulgation en évaluant dans quelle mesure la communication d’information sur le capital social et le capital humain réduit l’asymétrie informationnelle entre les dirigeants et les investisseurs. L’étude montre comment le marché boursier réagit aux différents niveaux de précision de l’information communiquée. Globale-ment, il ressort de l’étude que l’information quantitative réduit la volatilité des cours boursiers et accroît le ratio de Tobin (Q). Bien plus, la taille d’une entreprise modère la relation entre la précision de l’information et l’asymétrie informationnelle. Par ailleurs, pour déter-miner le niveau de précision de l’information divulguée, les entreprises prennent en compte les coûts et avantages ultimes des investisseurs. Enfi n, une gouvernance plus effi ciente se traduit par une plus grande transparence. Copyright © 2009 ASAC. Published by John Wiley & Sons, Ltd.

Mots-clés : asymétrie informationnelle, divulgation du capital social, divulgation du capital humain, gouvernance d’entreprise

SOCIAL AND HUMAN CAPITAL DISCLOSURE AND INFORMATION ASYMMETRY CORMIER ET AL.


Both human capital and social capital are considered key drivers of fi rm value in our knowledge-based society (Pennings, Lee, & Van Witteloostuijn, 1998). Human capital (HC) refers to the level of expertise and experi-ence of a fi rm’s employees (Blair & Kochan, 2000). Social capital (SC) refers to networks, social norms, and other relationships, either within or outside the fi rm, that allow it to operate and thrive (Cohen & Prusak, 2001). We assessed whether the determination of SC and HC disclosure and information asymmetry are closely inter-twined processes. Specifi cally, we set out to assess how the precision attribute (indicative, qualitative, quantita-tive) of a fi rm’s social disclosure affects information asymmetry between managers and investors. We defi ned indicative disclosure as “soft” or unverifi able disclosure. Our measure of social disclosure relates to a fi rm’s social capital and human capital and our proxies for informa-tion asymmetry between management and investors are the fi rm’s share price volatility and Tobin’s Q. Specifi -cally, consistent with the view that the determination of SC and HC disclosure and information asymmetry are closely intertwined processes, our research design takes into account the simultaneous effects of a fi rm’s disclo-sure strategy on: (a) the determinants of precision attri-bute of social and human capital disclosure and, (b) the impact of precision attribute of social and human capital disclosure on information asymmetry.

This paper builds on two prior studies that focused on the usefulness and relevance of corporate performance disclosure for market participants. Aerts, Cormier, and Magnan (2007) performed a comparative study of how fi nancial analysts use corporate performance disclosure by European and North American fi rms in their earnings forecasting work. Cormier, Ledoux, and Magnan (2009) showed that, for a Canadian sample, social-related dis-closure enhances a fi rm’s earnings valuation multiple. The current study focuses on how corporate disclosure about its social capital and human capital affects infor-mation asymmetry between managers and investors. Moreover, in contrast to prior work that relied on a single disclosure score, we assessed the differential impact of quantifi ed versus qualitative information dis-closure on information asymmetry. Although quantifi ca-tion is central to accounting, little is known about its effectiveness on reducing information asymmetry. Finally, the current study integrates fi rm-specifi c gover-nance and monitoring as potential determinants of disclosure.

Consistent with Clarkson, Li, Richardson, and Vasrini (2008), we focused on web-based disclosure. The Web has led fi rms to adapt their disclosure strategies since it offers much more fl exibility than traditional external reporting means. The Internet is widely seen as the best

platform for the disclosure management and stewardship of fi nancial and nonfi nancial information (Marston & Polei, 2004; Robb, Single, & Zarzeski, 2001). Moreover, there is essentially no marginal distribution cost if addi-tional information is released.

Our results from simultaneous equations indicate that quantitative verifi able (hard) disclosure is associated with less share price volatility and with a higher Tobin’s Q for social capital disclosure as well as for human capital disclosure. However, indicative unverifi able (soft) disclosure is marginally associated with a reduction in share price volatility and an increase in Tobin’s Q for human capital disclosure. As expected, fi rm size attenu-ates the impact of disclosure on information asymmetry. Furthermore, consistent with the nature of their disclo-sure, it does appear that fi rms take into account ultimate costs and benefi ts to shareholders when determining the extent of their disclosure. Variables that proxy for infor-mation costs are generally associated with quantitative disclosure. We also documented that effi cient governance leads to more transparency in quantitative human capital disclosure while the extent of CEO option pay leads to less transparency in social capital disclosure, either soft or hard.

Our contribution to prior research is twofold. Firstly, we assessed the informational properties of different dis-closure elements according to their precision attribute, that is, indicative, qualitative, or quantitative. Prior research has typically assumed that quantitative and monetary information is more informative than qualita-tive information (e.g., Aerts, Cormier, & Magnan, 2008; Al-Tuwaijri, Christensen, & Hughes II, 2004; Cho & Patten, 2007). However, that assumption has never been formally tested. Hence, we substantiated the differential effect of information precision in its effect on informa-tion asymmetry. Secondly, to the best of our knowledge, this study is the fi rst to investigate the impact of the nature of social and human capital disclosure on informa-tion asymmetry.

Benefi ts and Costs of Disclosure

By reassuring a fi rm’s investors about 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 information costs to be incurred by investors (e.g., Kim & Verrecchia, 1994). Healy and Palepu (2001) illustrated the perceived importance of information asymmetry and hence, of information costs as determinants for corporate disclosure. Therefore, a fi rm may decide to voluntarily disclose information if doing so is less costly than having



investors incur information costs themselves (Atiase, 1985; Lang & Lundholm, 1993). This, in turn, brings benefi ts to a fi rm by allowing it to lower its cost of capital (Botosan, 1997; Sengupta, 1998), to raise its valuation multiples, to increase share liquidity, and to enhance interest by institutional investors (Healy, Hutton, & Palepu, 1999).

However, expanded disclosure may also lead a fi rm to incur costs that result from external parties with diver-gent interests using its proprietary information to take actions that reduce its future cash fl ows (Dye, 1985). Such costs are deemed proprietary (Scott, 1994; Verrec-chia, 1983). Hence, a fi rm’s decision to disclose informa-tion to investors must weigh two opposite effects. Firstly, such disclosure can damage their competitive position in product markets (Healy & Palepu, 2001; Verrecchia, 1990). Secondly, disclosure of proprietary information may enhance a fi rm’s reputation as a credible discloser (Healy et al., 1999; Skinner, 1994).

Why are Social and Human Capital Disclosures Relevant to Investors?

Social and human capital, which rely extensively on relationships between an organization and its employees, business partners, and other stakeholders, provide unique opportunities to create value (Burt, 1992). For instance, Adler and Kwon (2002) noted that social and human capital facilitates various important organizational actions such as interunit and interfi rm learning, thus contributing to their success. Moreover, in building social and human capital, a fi rm is able to effectively reduce its market-based risk profi le (Orlitzky & Benjamin, 2001). For example, Waddock and Graves (1997) suggested that stable relations with various stakeholder groups facilitate a fi rm’s access to equity markets. Improvements in social and human capital also build trust in contracting relation-ships with external stakeholders, thus enabling the fi rm to lower transaction costs (Hill, 1990) and subsequent monitoring and coordination costs (Milgrom & Roberts, 1992). Hitt, Lee, and Yucel (2002) have shown that mul-tinational fi rms with extensive social capital exhibit a competitive advantage in the new global marketplace. In contrast, fi rms with limited relationship networks face high opportunity costs in their drive to expand internationally.

Shane and Cable (2002) have shown how direct and indirect network ties between entrepreneurs and potential investors, especially venture capitalists, bridge the infor-mation gap and facilitate the fi nancing of new ventures. However, the impact of a fi rm’s social and human capital disclosures on information asymmetry between manag-

ers and investors can only be effective if the fi rm’s social and human capital traits are visible and salient in the market, for example through social performance reputa-tion ratings (Fombrun & Shanley, 1990). In this vein, a corporate disclosure policy is important in supporting lasting effects of its social and human capital on market-based risk and performance measures. In that regard, Cormier et al. (2009) showed that social and human capital disclosure reduces a fi rm’s cost of equity capital.

In contrast to corporate social performance reputa-tion ratings, which are mainly sourced by well-respected and trustworthy third parties (e.g., Fortune ratings), dis-cretionary corporate disclosures may suffer from a credibility defi ciency, given management’s incentives to report in an opportunistic way. More generally, Hirst, Koonce, and Simco (1995) have shown that fi nancial reporting users consider the consistency between man-agement’s reporting incentives and a fi rm’s disclosures in assessing reporting credibility. This tendency increases with the level of discretion in reporting, but it decreases with a fi rm’s reporting reputation (Hodge, Hopkins, & Pratt, 2006). In this vein, fi rms may play on information attributes, like precision level, to mitigate an external audience’s tendency to discount the credibility of its dis-cretionary disclosures and thus, enhance the persuasive-ness and effectiveness of social and human capital disclosures.

Attributes of Disclosure, Disclosure Credibility, and Information Asymmetry

Mercer (2004) argued that information precision is one of the information attributes underlying disclosure credibility. Disclosures vary in degree of precision, ranging from indicative and qualitative information to specifi cally quantifi ed and monetary content. One of the reasons why quantitative information is more credible than soft, qualitative information is that it increases the ex post verifi ability of the information disclosed (Hutton, Miller, & Skinner, 2003). Moreover, quantitative disclo-sure may be viewed by managers as having proprietary value. Compared to qualitative disclosure, proprietary costs resulting from disclosure are likely to be higher for hard disclosure than soft disclosure since the fi rm may be perceived to reveal more credible and precise infor-mation (Cho & Patten, 2007). To the extent that quantita-tive information contains proprietary information, investors, knowing that disclosing proprietary informa-tion is costly, may perceive quantitative information to be more credible. In addition, quantifi ed information is likely to be normative in the sense that it allows



comparability through time or in space. Given this nor-mative potential, it may signal suffi ciency of corporate actions and responses (Anderson, Kadous, & Koonce, 2004) and as such be more persuasive than qualitative information.

The distinction between quantitative and qualitative information, and their underlying credibility, has funda-mental implications for information asymmetry. For example, fi rms shamed into disclosure by negative media coverage may tend to give out more soft, hard-to-verify, environmental information. This is known as “green washing,” and in so doing, fi rms may state they have an environmental policy rather than providing hard infor-mation such as benchmarking their performance to industry averages (Clarkson et al., 2008). Such informa-tion is likely to have a low impact, if any, on the reduc-tion of uncertainty for investors. Furthermore, we can argue that hard disclosure is more diffi cult to mimic by a fi rm’s competitors than soft disclosure, thus enhancing its credibility. Quantitative disclosure may be viewed by managers as having proprietary value. Proprietary costs resulting from disclosure are likely to be higher for hard disclosure than soft disclosure since the fi rm may be perceived to reveal more credible and precise informa-tion (Cho & Patten, 2007).

Hypotheses

Firms with high share price volatility have incen-tives to reduce information asymmetry between manag-ers and investors since such actions lower fi nancing (Clarkson, Kao, & Richardson, 1994; Frankel, Mc-Nichols, & Wilson, 1995). By reassuring a fi rm’s inves-tors 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 information costs incurred by investors (e.g., Kim & Verrecchia, 1994). Several approaches to assess information asymmetry have been proposed. Francis, Khurana, and Pereira (2005), Leuz and Verrecchia (2000), Healy et al. (1999), and Welker (1995) have all shown that the extent of information asymmetry—proxied by bid-ask spread, share price volatility, or share liquidity—is negatively associated with disclosure quality.

With respect to social disclosure, Richardson and Welker (2001) found that, contrary to expectations, more disclosure translates into higher cost of capital. One potential explanation for this fi nding is that they did not control for endogeneity between disclosure and fi rm attributes. Cormier et al. (2009) showed that a fi rm’s web-based disclosure affects earnings valuation

multiples, but in a differential way according to the nature of the information disclosed. More specifi cally, disclosure about human capital (and to a lesser extent social capital) is associated with a larger earnings multi-ple (e.g., a lower cost of capital), but this is not the case for disclosure related to value for clients or corporate governance.

Consistent with Clarkson et al. (2008), our chosen proxies for information asymmetry are share price volatility and Tobin’s Q. We expected a lower negative relationship between soft disclosure (indicative or qualitative–unverifi able disclosure) and information asym-metry than for hard (quantitative–verifi able) disclosure. Hence:

H1) The negative association between information asymmetry and the extent of voluntary disclosure about human capital and social capital will be higher for quantitative disclosure than for indicative and qualitative disclosure.

Prior research has documented that information asymmetry is lower for large fi rms compared with small fi rms (Leuz & Verrecchia, 2000). Size controls for a fi rm’s information environment and reporting reputation. The number of investors in large fi rms is signifi cantly higher than in small fi rms. Therefore, the amount of liquidity trading should be signifi cantly higher for large fi rms than for small fi rms. Cheng, Gopinath, and Krish-namurti (2002) showed that an increase in trading activity of large fi rms leads to improvement in liquidity while an increase in small fi rms’ trading activity leads to deterioration in liquidity. They documented that while more frequent trading is associated with an improvement in liquidity, as proxied by the bid-ask spread for large market capitalization stocks, the opposite is true for small stocks.

Therefore, we expected disclosure to have a lower impact on share price volatility for large fi rms since investors have access to more information. Usually, large fi rms are highly followed by analysts. The effect of dis-closure on properties of analysts forecast dispersion and accuracy is signifi cantly different for small fi rms versus large fi rms. The disclosure effect is more marked for small fi rms with a low analyst following than for large fi rms with a high analyst following (Botosan, 1997). Moreover, Hodge et al. (2006) suggested that the external users’ tendency to discount the credibility of discretion-ary disclosures decreases with a fi rm’s reporting reputa-tion, which would erode the differential impact of information precision on information asymmetry for larger fi rms. Thus, we expected smaller fi rms to have lower reporting reputations. Hence:



H2) The negative association between information asymmetry and the extent of voluntary disclosure about human capital and social capital will be lower for large fi rms than for small fi rms.

Method

Sample

The sample comprised 155 observations of web dis-closure for the year 2005. We initially collected web disclosure in the summer of 2002 for an international study (Aerts et al., 2007). All nonfi nancial fi rms repre-sented on the Toronto Stock Exchange S&P/TSX Index were identifi ed. To ensure that our sample comprised fi rms with active information dynamics and investor interest, we selected fi rms that were followed by at least two fi nancial analysts. The resulting 2002 sample com-prised 189 nonfi nancial fi rms. Mergers and acquisitions as well as bankruptcies and delistings reduced our sample to 155 in 2005. The sample fi rms represent more than 55% of the Toronto Stock Exchange stock market capi-talization for nonfi nancial fi rms. The same coding grid has been used by Aerts et al. (2007) and Cormier et al. (2009).

Financial data for 2004 were collected from Stock Guide (2005) and governance data were collected from 2004 proxy statements. The fi nal sample comprised 131 fi rms since, out of the initial sample of 155 fi rms, there were missing data for board size and board independence (2 fi rms), stock options (12 fi rms), and share volatility (12 fi rms). Sample fi rms operated in the following indus-tries: metals and mines; oil and gas; paper and forest products; consumer products; industrial products; real estate; utilities; communication and media; and merchandising.

Empirical Model

We attempted to provide an integrated analysis of fi rms’ web-based disclosure strategy. We posited that this strategy simultaneously affects share price volatil-ity/Tobin’s Q and disclosure. It is thus important to actu-ally control for the presence of endogeneity between our critical variables. Endogeneity is of concern if a disclo-sure decision is made simultaneously to share price vola-tility/Tobin’s Q, and if we fail to take into consideration the factors that explain the disclosure decision. Given the endogenous nature of disclosure and our research hypoth-eses, we built the following two systems of structural equations. Note that while the explanatory variables for volatility and Tobin’s Q are contemporaneous, the

explanatory variables for disclosure are lagged, as explained in the section entitled “Multivariate Analyses- Share Price Volatility.”

(1a) Share price volatility it = ƒ( 0 + 1Size + 2Systematic risk + 3Free fl oat + 4Indicative Disclosure + 5Qualitative Disclosure + 6Quantitative Disclosure + 7Indicative Disclosure*Size + 8

Qualitative Disclosure*Size + 9Quantitative Disclosure*Size)it

(1b, c and d) Disclosure itk = ƒ( 0 + 1New Financing + 2Free fl oat + 3Analyst following + 4Leverage + 5Profi tability + 6US listing + 7Size + 8Board

independence + 9Board size + 10Audit committee size + 11CEO stock options)it-1

(2a) Tobin’s Q it = ƒ( 0 + 1Inverse of assets + 2Return on assets + 3Indicative Disclosure + 4Qualitative Disclosure + 5Quantitative Disclosure)it

(2b, c and d) Disclosure itk = ƒ( 0 + 1New Financing + 2Free fl oat + 3Analyst following + 4Leverage + 5Profi tability + 6US listing + 7Size + 8Board

independence + 9Board size + 10Audit committee size + 11CEO stock options)it-1

Where i is a specifi c fi rm, t is a given year, and k represents a disclosure precision attribute, that is, indica-tive, qualitative, or quantitative. Hence, there are three estimated equations for disclosure. Within our model, we integrated the three disclosure levels simultaneously. By doing so, we implicitly controlled for the relative effect of the different information precision levels on share price volatility/Tobin’s Q.

Measurement of Information Asymmetry

Share price volatility is defi ned as the standard devi-ation of percentage changes in daily stock prices for 2005. Tobin’s Q is defi ned as the market value of a fi rm’s equity, plus the book value of its debt, divided by the book value of equity and debt at the end of 2005.

Measurement of Web-Based Disclosure

Performance disclosure indicators, fi nancial or non-fi nancial, are based on balance scorecard literature and emerging performance measurement practices (e.g., for fi nancial and governance disclosure [Standard & Poors, 2002]; for investors, governance, and social responsibil-ity disclosures [Marston & Polei, 2004; Pirchegger & Wagenhofer, 1999]; and for indicators about operations’ effi ciency, value for client, innovation, development, and growth [Ittner & Larker, 1998; Kaplan & Norton, 1996; Robb et al., 2001]). We measured disclosure using a



coding instrument that is consistent with Cormier et al. (2009), Aerts et al. (2007), Cormier and Magnan (2003), and Wiseman (1982).

The grid comprises 33 items. The 33 disclosure items are grouped into two categories: human capital and social capital. In a manner consistent with prior litera-ture, human capital items relate to activities, actions, or outcomes that refl ect employees’ expertise and experi-ence (Pennings et al., 1998). Social capital refers to fea-tures of social organization such as networks, norms and social trust that facilitate co-operation for mutual advantage (Industry Canada, 2008). Social capital items relate to interactions between the fi rm and society (e.g., alliances, clients) and within the fi rm itself (i.e., with employees) (e.g., Dess & Shaw, 2001; Pastoriza, Arino, & Ricart, 2008). We provide details of our categorization of social and human capital in the Appendix.

The web content is split among qualitative elements (indicative, descriptive) and monetary or quantitative elements. Each element is scored 1, and an item may include many elements. Our disclosure measure com-prised only information on a fi rm’s website provided in an HTML format and specifi cally excluded mandated corporate documents that were linked to the website. Websites were analyzed and coded online with the website information being kept on a CR-Rom. To ensure consistency across fi rms, two persons reviewed indi-vidual scores independently. All disagreements were reviewed by one of the coresearchers.1

We focused on voluntary disclosure on a fi rm’s website, as it is comprehensive and accessible to every-one at a low cost. Our disclosure measure comprised only information on a fi rm’s website and did not encompass paper-based disclosure. In a Canadian context, mandated disclosure is being fi led on SEDAR (a system of elec-tronic data archiving and retrieval that is maintained by securities regulators). SEDAR includes all documents where disclosure is mandated by securities regulators: fi nancial statements, annual reports, proxy statements, Management Discussion and Analysis (MD&A), and press releases (which typically concern material changes). However, all of these documents are available in paper form. The content of all of these documents is prespeci-fi ed and regulated. Therefore, our disclosure measure is likely to refl ect mostly voluntary information.

The use of a coding scale to qualify a fi rm’s disclo-sure is appropriate for the following reasons. Firstly, it allows for an integration of different types of information into a single fi gure that is comparable across fi rms in terms of relevance. Secondly, a qualitative scale allows the researcher’s judgment to be impounded in rating the value or quality of the disclosure made by a fi rm. While this process is more subjective, it ensures that irrelevant

or redundant generalities are not considered strategic disclosure.

Measurement of Explanatory Variables

Determinants of information asymmetry (share price volatility/Tobin’s Q). Prior studies on the determi-nants of the information asymmetry component of cost of capital suggest numerous determinants other than voluntary disclosure (Leuz & Verrecchia, 2000). Based on that literature, we used fi rm size, systematic risk, free fl oat, and analyst following as determinants of informa-tion asymmetry.

Size. Several studies (e.g., Botosan, 1997; Botosan & Plumlee, 2005; Gebhardt, Lee, & Swaminathan, 2001; Sengupta, 1998) found that larger fi rms benefi t from a lower information asymmetry and therefore a lower cost of capital. We thus expected information asymmetry to be negatively associated with fi rm size.

Systematic risk. The higher a fi rm’s systematic risk, the more diffi cult it is for investors to precisely assess a fi rm’s value and the more likely they are expected to incur information costs to assess its risk drivers. Inves-tors charge a higher cost of equity for fi rms with higher systematic risk (e.g., Botosan & Plumlee, 2005; Hail & Leuz, 2006; Gebhardt et al., 2001; Leuz & Verrecchia, 2000). We expected a positive relationship between sys-tematic risk and information asymmetry.

Free fl oat. We used free fl oat as an inverse proxy for the presence of insiders since control blocks generally have greater access to private information (Leuz & Verrecchia, 2000). Hence, we expected a nega-tive association between free fl oat and information asymmetry.

Determinants of Web-Based Disclosure

Verrecchia (1983) argued that whether a fi rm will voluntarily disclose corporate information is a function of the proprietary costs associated with the disclosure. Unless there is perceived benefi t that outweighs the pro-prietary cost, fi rms will not disclose. We could argue that quantitative disclosures are likely to be viewed by management as having higher proprietary value (they will reveal more to competitors, for example) than quali-tative disclosures, and as such, will be less preferred as a disclosure tool than other types of disclosure, except when the contracting incentives (e.g., new fi nancing) or governance pressures are high enough to induce the dis-closure of information potentially more litigious (quan-titative versus qualitative disclosure).



Proxies for Information Costs

Three variables are used to capture investors’ infor-mation needs and information costs with respect to a fi rm’s web-based disclosure: new fi nancing, free fl oat, and analyst following.

New fi nancing. Lang and Lundholm (1993) docu-mented a positive relationship between the need for fi nancing and voluntary disclosure (as measured by fi nancial analysts’ disclosure scores). The variable measures long term fi nancing raised through share or debt offerings scaled by total assets. We expected a posi-tive relationship between the variable new fi nancing and web-based disclosure.

Free fl oat. Ownership structure can determine the level of monitoring and, thereby, the extent of disclosure (Eng & Mak, 2003). Firms with widely-held ownership are expected to be responsive to public investors’ infor-mation costs since no dominant shareholders typically have access to the information they need (Hope, 2003). Yhim, Karin, and Rutledge (2003) showed that the pro-portion of outside ownership is signifi cantly associated with high-level forecast precision. A positive relation was therefore expected between free fl oat and disclosure.

Analyst following. Lang and Lundholm (1996) and Healy et al. (1999) found a positive relationship between analyst following and the quality of a fi rm’s disclosure. Hence, we expected a positive relationship between analyst following and the extent of disclosure.

Proxies for Litigation/Proprietary Costs

A priori, the magnitude of potential costs a fi rm faces because of disclosure is diffi cult to assess since it requires the identifi cation of all parties that may use information to the fi rm’s detriment. However, a fi rm’s fi nancial condition does provide a measure of its willing-ness to release proprietary information since only fi rms that are fi nancially sound may be able to trade off the benefi ts from additional disclosure with the costs of revealing potentially damaging information. In contrast, fi rms in poor fi nancial condition may be unable to with-stand the initial negative consequences that are needed to gain any benefi ts from more extensive disclosure. Four variables proxy for a fi rm’s ability to support proprietary costs: profi tability, leverage, US listing, and fi rm size.

Profi tability. Many studies document a positive association between a fi rm’s level of disclosure and its fi nancial performance (Cormier & Magnan, 2003; McGuire, Sundgren, & Schneeweis, 1988). A positive relationship is expected between profi tability, free fl oat, and disclosure.

Leverage. Firms in poor fi nancial condition may not be able to withstand the initial negative consequences that are needed to gain any benefi ts from more extensive disclosure. Thus, consistent with prior fi ndings (Cormier & Magnan, 2003; McGuire et al., 1988), we expected a negative relationship between a fi rm’s leverage and disclosure.

Firm size. Prior evidence is consistent in showing a positive relationship between the extent of corporate disclosure and fi rm size (Neu, Warsame, & Pedwell, 1998; Scott, 1994). Firm size also proxies for other factors, such as the extent of monitoring by analysts. Firm size is introduced with an expectation of a positive relation with disclosure.

Proxies for Governance Monitoring

Prior studies have shown that a fi rm’s governance infl uences the quality of its voluntary disclosure (Eng & Mak, 2003). More specifi cally, the intensity of monitor-ing by a board has a direct infl uence on managerial dis-cretion and typically requires fi rms to engage in more extensive organizational performance measurement and reporting (Fama, 1980). Such monitoring by the board can be implemented through various means and attri-butes. For instance, an independent board is more likely to be effective in assessing managerial decisions and performance than a board that comprises only insiders (Beasley, 1996; Fama & Jensen, 1983; Xie, Davidson, & DaDalt, 2003). With respect to voluntary disclosure, Chen and Jaggi (2000) documented that a board that comprises mostly independent nonexecutive directors is more likely to be associated with comprehensive fi nan-cial disclosure. Moreover, Karamanou and Vafeas (2005) showed that fi rms with better governance are more likely to issue voluntary earnings forecasts. However, Cheng and Courtenay (2005) found that the relation between governance and disclosure is enhanced if there is an effi cient regulatory environment.

Furthermore, stock options can align manager inter-ests with shareholder interests (Hanlon, Rajgopal, & Shevlin, 2003). However, contracting costs may lead to incomplete contracts and agency confl icts. Aboody and Kaznik (2000) showed that managers with stock-based compensation mislead shareholders by accelerating bad news and by infl uencing the timing of news. Hence, web-based disclosure is likely to be opportunistically affected by CEO stock options.

Four variables were introduced to capture the impact of corporate governance as a monitoring factor affecting web-based disclosure.

Board independence. We expected board indepen-dence, measured as the proportion of outside directors,



to be positively associated with voluntary disclosure. Another aspect of board independence is the separation of the roles of Chair and Chief Executive Offi cer. Our variable takes the value of zero (0) when the majority of directors are not independent, one (1) when the majority of directors are independent, and two (2) when the major-ity of directors are independent and the function of CEO and Chair of the board is separate. We expected a positive relationship between this variable and disclosure.

Board size and audit committee size. Beasley (1996) found a positive relationship between board size and the likelihood of fi nancial statement fraud while Abbott, Park, and Parker (2000) found no relationship. Moreover, Bédard, Chtourou, and Courteau (2004) found that board size was associated with less earnings management but only for income decreasing accruals. Additionally, we believe that audit committee effectiveness is a critical determinant of voluntary web-based disclosure. In Canada, audit committees must comprise at least three independent members, which is a small number if the audit committee is to be effective in its monitoring role. Since adding a few members to the audit committee could enhance this monitoring role, we expected board size and audit committee size to relate positively to disclosure.

Stock options. The importance of contracting costs may lead to incomplete contracts and agency confl icts. The more agency confl icts between managers and share-holders are important, the more managers with stock-based compensation will manage disclosure to maximize

the value of their stock options. We expected disclosure to be opportunistically affected by the presence of CEO stock options. Since the actual impact of stock options on reporting is unclear, no directional predictions were made. In Table 1, we defi ne how the independent vari-ables of our study were measured.

Results

Descriptive Statistics

Table 2a provides descriptive statistics on fi nancial and governance variables for the sample fi rms. About 78% of sample fi rms have a diffused ownership and 8.6% of sample fi rms had recently relied on capital markets for additional fi nancing. CEO value of in-the-money exer-cisable stock options represented almost twice their salary and bonus (1.79). On average, 36% of directors are independent, and 20% of CEOs also hold the position of chairman of the board. Table 2b provides correlations between our fi nancial and governance variables. The cor-relation matrix did not show high correlations among independent variables. Only two of the correlation coef-fi cients exceeded 0.50 (board size and audit committee size at 0.567 and fi rm size and board size at 0.53).

As illustrated in Table 3, on average, soft disclosure was larger than hard disclosure, both for social capital disclosure (16.05 items for indicative, 3.72 for qualita-tive, and 1.83 for quantitative) and human capital disclo-sure (3.83 items for indicative, 2.02 items for qualitative,

Table 1Measure of Independent Variables

Information costsSystematic risk Beta in 2004New fi nancing Long-term debt borrowing plus stock issued in 2004 scaled by total assetsFree fl oat The percentage of shares that are not closely held (total shares outstanding minus control

blocks of 10% or more)Analyst following Number of analysts following a fi rm in 2004

Litigation/Proprietary costsLeverage Long term debt / Total assetsProfi tability Return on assetsForeign listing SEC registration (binary variable 1; 0 if not)Firm size Ln(Total Assets) as of year-end 2004

Governance/MonitoringBoard independence (0) if a majority of directors are not independent; (1) if a majority of directors are independent;

(2) if a majority of directors are independent and if the function of CEO and Chair is separated

Board size Number of directors on the boardAudit committee size Number of audit committee membersCEO stock options Value of in-the-money exercisable stock options/salaries + bonus



and 0.68 item for quantitative). Internal consistency esti-mates (Cronbach’s alpha on score components) showed that the variance was quite systematic (alpha = 0.65 for human capital disclosure and 0.74 for social capital dis-closure). This is comparable to Botosan (1997) who found an alpha of 0.64 for an index including fi ve catego-ries of disclosure in annual reports. According to Nunnaly (1978), an alpha of 0.70 is acceptable.

Multivariate Analyses- Share Price Volatility

Since we posited that a fi rm’s information dynamics affect disclosure and share price volatility simultane-ously, we fi rst assessed whether or not endogeneity exists between these variables using a Hausman test. Using this procedure, we rejected the null hypothesis of no endoge-neity with respect to share price volatility and social capital disclosure (indicative: p < 0.07 [marginally], qualitative: p < 0.01, quantitative: p < 0.01), and share price volatility and human capital disclosure (indicative: p < 0.01, qualitative: p < 0.01, quantitative: p < 0.01). Therefore, share price volatility and disclosure variables were treated as endogeneous variables. In light of this diagnostic, we relied on a three-stage estimation framework (three-stage-least-square or 3SLS). A 3SLS combines two-stage-least-squares (2SLS) and seemingly-unrelated-regression estimation (SURE).

Within such an approach, all variables that proxied for information asymmetry and disclosure were treated as endogenous, but cross-equation error correlations improved sample effi ciency. The fi rst two stages of 3SLS were the same as 2SLS: in other words, the instrumental variables were used as regressors to obtain the projected value of the endogenous variables, and the projected values (rather than the original values) of the endogenous variables were used in the second stage regressions. In our case, we used the independent variables in equations 1 (b, c, and d) as instruments to obtain the fi tted values of our three disclosure variables and then substituted the fi tted values into equation 1(a) in lieu of the disclosure variables. Use of the lagged form of the explanatory variables for the disclosure regressions mitigated con-cerns about endogeneity and made these variables appro-priate to use as instruments (see Xue, 2007; Pyndick & Rubinfeld, 1991 for further details). In the third and fi nal stage, the 2SLS residuals were used to estimate the cross-equation error covariance matrix and to generate correlation coeffi cients, which was more effi cient than 2SLS.

Table 4 reports results of a three-stage estimation model for share price volatility, SC indicative disclosure, SC qualitative disclosure, and SC quantitative disclosure. Consistent with hypothesis 1, the association between share price volatility and the extent of SC disclosure was

Table 2aDescriptive Statistics: Financial and Governance Variables

Mean Median Std dev. Minimum Maximum

Information asymmetryShare price volatility (std dev. of percentage

change in daily stock prices for year 2005)2.233 1.809 1.494 0.818 10.385

Information costsSystematic risk 0.682 0.560 0.489 0 2.71New fi nancing 0.086 0.040 0.119 0 0.70Free fl oat 0.776 0.845 0.225 0.098 0.999Analyst following 6.829 6.000 5.888 0 35Litigation/Proprietary costsLeverage 0.220 0.200 0.214 0 2.00Profi tability 0.035 0.040 0.131 −1.07 0.56US listing 0.511 1.000 0.501 0 1Firm size (total assets in million Can $) 4,844 1,821 7,226 26 40,076Governance/MonitoringBoard independence (0,1,2) 0.909 1 0.515 0 2 Independent directors (%) 0.360 0.375 0.178 0 0.860 Board chair duality (%) 0.200 0 0.401 0 1Board size 9.987 10.000 2.755 4 18Audit committee size 3.980 4.000 1.103 2 9CEO stock options 1.786 1.785 21.715 0 229



Tab

le 2

bC

orre

latio

n M

atrix

: F

inan

cial

and

Gov

erna

nce

Var

iabl

es

Shar

e pr

ice

vola

tility

Syst

emat

icri

skN

ewfi n

anci

ngFr

eefl o

atA

naly

stfo

llow

ing

Lev

erag

ePr

ofi ta

bilit

yU

Slis

ting

Firm

size

Boa

rdin

depe

nden

ceB

oard

size

Aud

itco

mm

ittee

siz

eC

EO

sto

ckop

tions

Shar

e pr

ice

vola

tility

1Sy

stem

atic

ris

k*0

.278

1N

ew fi

nan

cing

*0.1

520.

046

1Fr

ee fl

oat

0.11

3*0

.168

0.11

81

Ana

lyst

fol

low

ing

*−0.

155

*0.3

27*−

0.17

70.

068

1L

ever

age

*−0.

204

−0.1

67*0

.158

*−0.

197

*−0.

209

1Pr

ofi ta

bilit

y*−

0.26

2−0

.046

*−0.

133

−0.0

340.

084

−0.0

091

US

listin

g0.

109

0.24

2*0

.175

*0.1

660.

092

0.10

4−0

.042

1Fi

rm s

ize

*−0.

494

−0.0

17−0

.095

*−0.

149

*0.2

07*0

.381

*0.1

460.

114

1B

oard

ind

epen

denc

e−0

.042

−0.0

07−0

.096

−0.0

64−0

.034

*−0.

178

−0.1

21*−

0.19

4−0

.065

1B

oard

siz

e*−

0.37

9*−

0.16

90.

015

−0.1

540.

073

0.19

30.

119

−0.1

07*0

.530

0.07

81

Aud

it co

mm

ittee

siz

e*−

0.31

7−0

.120

0.09

90.

015

−0.0

520.

129

0.13

4−0

.041

*0.4

090.

123

*0.5

671

CE

O s

tock

opt

ions

0.02

4*0

.221

−0.0

550.

037

*0.2

87−0

.123

0.04

1*0

.139

0.09

6−0

.006

−0.0

78−0

.002

1So

cial

cap

ital-

indi

cativ

e*−

0.24

1*−

0.07

5−0

.084

0.04

1*−

0.16

00.

031

*0.0

89*0

.172

*0.4

220.

016

*0.3

62*0

.362

−0.0

62So

cial

cap

ital-

qual

itativ

e*−

0.24

4*−

0.11

50.

012

0.07

1*0

.111

0.04

80.

077

*0.2

25*0

.493

−0.0

09*0

.407

*0.4

06−0

.041

Soci

al c

apita

l- q

uant

itativ

e*−

0.24

0*−

0.10

6*−

0.13

2−0

.037

0.08

0−0

.015

0.08

4*0

.139

*0.3

970.

012

*0.3

34*0

.329

−0.0

71H

uman

cap

ital-

indi

cativ

e−0

.133

*0.2

03*−

0.13

9−0

.085

*0.2

80−0

.043

−0.0

02*0

.109

*0.2

55−0

.097

*0.2

77*0

.180

0.04

8H

uman

cap

ital-

qual

itativ

e−0

.134

−0.1

33−0

.012

0.00

9*0

.177

−0.0

330.

014

*0.1

27*0

.312

0.04

2*0

.266

*0.1

80−0

.005

Hum

an c

apita

l-qu

antit

ativ

e−0

.106

−0.1

06−0

.066

0.03

40.

068

0.06

1−0

.072

0.05

8*0

.146

*0.1

49*0

.194

*0.1

650.

088

Not

e: *

: p

< 0

.10.



present only for quantitative disclosure (−5.069; p < 0.01). As expected (hypothesis 2), this association was reduced by fi rm size (0.225; p < 0.01). Furthermore, chi-square test for coeffi cient difference confi rms our results

since the sum of coeffi cients for Disclosure Quantitative and Disclosure quantitative*size was statistically close to zero ( 6 − 9 = 0; 4.86; p < 0.03), that is, for large fi rms both coeffi cients have about the same weight—one

Table 3Web-based Disclosure: Mean (Median) Scores by Component

2005

Indicative content Qualitative content Quantitative content Cronbach Alpha

Social capital 16.05 (10.00) 3.72 (2.00) 1.83 (1.00) 0.74Human Capital 3.83 (2.00) 2.02 (2.00) 0.68 (1.00) 0.65Total score 19.88 (14.00) 5.74 (4.00) 2.51 (2.00)N: 155

Table 4Three Stage Regression Model: (Seemingly Unrelated Regression Estimation) Social Capital Disclosure

N: 131 Share pricevolatility

Disclosureindicative

Disclosurequalitative

Disclosurequantitative

Share price volatilityFirm size − 1 ***−0.668Systematic risk + 2 ***0.583Free fl oat − 3 0.222Disclosure indicative − 4 0.614Disclosure qualitative − 5 −1.434Disclosure quantitative − 6 ***−5.069Disclosure indicative*size + 7 −0.026Disclosure qualitative*size + 8 0.064Disclosure quantitative*size + 9 **0.225DisclosureInformation costsNew fi nancing + −6.525 −0.606 −1.633Free fl oat + 3.774 **2.548 −0.376Analyst following + **0.326 −0.018 0.007Litigation/Proprietary costs ***−2.734Leverage − ***−17.468 *−2.658Profi tability + −2.260 −0.834 0.563Foreign listing + ***7.667 ***2.256 **0.544Firm size + ***5.127 ***1.239 ***0.541Governance monitoringBoard independence + 1.226 0.210 −0.022Board size + 0.301 *0.178 0.052Audit committee size + **2.269 **0.651 **0.302CEO stock options +/− ***−0.138 *−0.023 **−0.014Adjusted R2 40.3% 45.9% 43.5% 28.8%Chi2 88.6 111.0 100.8 52.9P value 0.000 0.000 0.000 0.000Test 6 − 9 = 0 Chi2-test 4.86(0.027)

Note: *: p < 0.10; **: p < 0.05; ***: p < 0.01. One-tailed if there is a predicted sign, two-tailed otherwise.



in a positive direction and the other in a negative direc-tion. Therefore, for large fi rms, there was no signifi cant relationship between SC quantitative disclosure and share price volatility.

Concerning the determinants of SC disclosure, liti-gation/proprietary costs (negative association for lever-age, positive association for foreign listing and fi rm size) and corporate governance (positive association for audit committee size) are associated with disclosure precision, that is, indicative, qualitative, or quantitative. The mag-nitude of the CEO’s stock option value was negatively associated with all SC disclosure. These results suggest that effi cient governance leads to more transparency while the extent of CEO stock options lead to less transparency. The fact that voluntary disclosure could

be opportunistically affected by the extent of the CEO stock option is consistent with Aboody and Kaznik (2000).

Table 5 reports results of a three-stage estimation model for share price volatility, HC indicative disclosure, HC qualitative disclosure, and HC quantitative dis-closure. Consistent with hypothesis 1, the association between share price volatility and the extent of HC dis-closure was present for quantitative disclosure (−3.343; p < 0.05). As expected (hypothesis 2), this association was reduced by fi rm size (0.155; p < 0.05). Furthermore, a chi-square test for coeffi cient difference confi rms our results since the sum of coeffi cients for Disclosure Quan-titative and Disclosure quantitative*size is statistically close to zero ( 6 − 9 = 0; 2.93; p < 0.08) that is, for large

Table 5Three Stage Regression Model: (Seemingly Unrelated Regression Estimation) Human Capital Disclosure





Share price volatilityFirm size − 1 ***−0.645Systematic risk + 2 ***0.566Free fl oat − 3 0.307Disclosure indicative − 4 *−0.615Disclosure qualitative − 5 0.018Disclosure quantitative − 6 **−3.433Disclosure indicative*size + 7 *0.028Disclosure qualitative*size + 8 −0.001Disclosure quantitative*size + 9 **0.155DisclosureInformation costsNew fi nancing + −3.497 1.012 0.257Free fl oat + −1.745 *1.087 0.213Analyst following + **0.098 *0.048 0.007Litigation/Proprietary costs **−0.984Leverage − −1.134 −1.484Profi tability + −0.263 −1.293 −0.134Foreign listing + **1.282 **0.779 **0.227Firm size + ***0.801 ***0.615 ***0.154Governance monitoringBoard independence + −0.258 0.295 ***0.364Board size + **0.252 *0.112 0.026Audit committee size + 0.048 −0.092 −0.002CEO stock options +/− −0.006 *−0.010 0.001Adjusted R2 36.6% 24.6% 25.6% 19.5%Chi2 75.7 42.8 45.2 31.7P value 0.000 0.000 0.000 0.000Test 4 − 6 = 0 Chi2-test 1.36 (0.243)

Test 4 − 7 = 0 Chi2-test 2.01 (0.155)

Test 6 − 9 = 0 Chi2-test 2.93 (0.087)




fi rms, there was no signifi cant relationship between HC quantitative disclosure and share price volatility.

There was also an association between indicative soft disclosure and share price volatility. Moreover, a chi-square test for coeffi cient difference confi rmed our results since the sum of coeffi cients for Disclosure- indicative and Disclosure-indicative*size was marginally statistically close to zero ( 4 − 7 = 0; 2.01; p < 0.15); that is, for large fi rms there was no signifi cant relation-ship between HC indicative disclosure and share price volatility. One explanation for the association between soft HC disclosure and share price volatility could be that some indicative disclosure components could be seen as relevant by market participants. In our sample, more than half of the fi rms disclosed at least one indicative item of information concerning employee competence.

Concerning the determinants of HC disclosure, we observed that foreign listing and fi rm size are associated with indicative, qualitative, and quantitative disclosure precision, while board independence is associated only with hard quantitative disclosure. The magnitude of the CEO’s option value was negatively associated with qual-itative HC disclosure.

Both social capital and human capital were closely related. This fi nding is intuitive, because without social capital, it may be diffi cult for a fi rm to attract and develop its human capital. Inversely, without human capital, it will be quite a challenge for a fi rm to build social capital as employees’ commitment and efforts will be lacking. Hence, as an additional analysis, we assessed the rela-tionship between combined social and human capital disclosure and share price volatility. Table 6 shows

Table 6Three Stage Regression Model: (Seemingly Unrelated Regression Estimation) Combined Human and Social Capital Disclosure





Share price volatilityFirm size − 1 ***−0.673Systematic risk + 2 ***0.593Free fl oat − 3 0.239Disclosure indicative − 4 −0.038Disclosure qualitative − 5 0.016Disclosure quantitative − 6 **−2.081Disclosure indicative*size + 7 0.002Disclosure qualitative*size + 8 −0.001Disclosure quantitative*size + 9 **0.093DisclosureInformation costsNew fi nancing + −10.129 0.396 −1.389Free fl oat + 2.036 **3.636 −0.162Analyst following + **0.441 0.032 0.016Litigation/Proprietary costs ***−3.691Leverage − ***−18.391 **−4.125Profi tability + −1.699 −2.056 0.537Foreign listing + ***8.886 ***3.028 **0.761Firm size + ***5.871 ***1.849 ***0.687Governance monitoringBoard independence + 0.926 0.502 0.337Board size + 0.555 **0.291 0.078Audit committee size + **2.411 *0.566 **0.311CEO stock options +/− ***−0.146 **−0.033 *−0.014Adjusted R2 39.1% 48.9% 46.9% 33.9%Chi2 84.8 126.5 115.8 67.7P value 0.000 0.000 0.000 0.000Test 6 − 9 = 0 Chi2-test 2.92 (0.089)




regression results for combined social and human capital disclosure. Consistent with our hypotheses, the results suggest that, overall, only quantitative disclosure reduces share price volatility.

To test whether our results could be affected by multicollinearity between interaction terms and main effect coeffi cients, we estimated an OLS regression on a subsample based on size. We considered small fi rms to be those included in the fi rst 25th percentile (natural log of assets smaller or equal to 20.37). Results presented in Table 7 support those presented in Table 4 and 5. These results are consistent with hypotheses 1 and 2.

Endogeneity concerns the problem of omitted vari-ables and of simultaneity. Our models controlled for omitted variables. Simultaneity arose when one or more of the explanatory variables were jointly determined with the explained variable (Chenhall & Moers, 2007). To test whether our results could be affected by the simultaneity between disclosure and share price volatility, we intro-duced share price volatility in disclosure regressions. Results (not tabulated) show that share price volatility did not signifi cantly determine social capital disclosure (indicative p < 0.28; qualitative disclosure p < 0.34; quantitative disclosure p < 0.97) while there was a posi-tive association for indicative and qualitative human capital disclosure (indicative p < 0.01; qualitative disclo-sure p < 0.05; quantitative disclosure p < 0.69). However, results for volatility determinants remained similar to those presented in Table 4 and Table 5.

Multivariate Analyses- Tobin’s Q

As a second proxy for information asymmetry, we used Tobin’s Q. Since Tobin refers to market to book

value of assets, we scaled our disclosures, variables by book value of assets. Again, since we posited that a fi rm’s information dynamics affect disclosure and share price volatility simultaneously (re: equations 2b, c, d), we fi rst assessed whether or not endogeneity exists between these variables using a Hausman test. Using this procedure, we could not reject the null hypothesis of no endogeneity with respect to Tobin’s Q and social capital disclosure (indicative: p < 0.66, qualitative: p < 0.37, quantitative: p < 0.22), and Tobin’s Q and human capital disclosure (indicative: p < 0.54, qualitative: p < 0.13, quantitative: p < 0.21). Therefore, we used Ordinary Least Square OLS regression estimations. Results presented in Table 8 are consistent with our hypotheses since quantitative disclosure scaled by assets related positively to Tobin’s Q (social capital: 1.543; p < 0.01 and human capital: 1.805; p < 0.05). Similar to the results for share price volatility, our results show an association between indic-ative soft disclosure and Tobin’s Q.

Discussion

Summary

We built on prior literature on voluntary disclosure by investigating the impact of precision attribute of dis-closure on information asymmetry in the stock market. Our results suggest that quantitative (hard) social and human capital disclosure reduces information asymme-try, as proxied by share price volatility and Tobin’s Q. Furthermore, indicative (soft) human capital disclosure is marginally associated with a reduction in information asymmetry. As expected, fi rm size reduces the impact of

Table 7Ordinary Least Square Regression Model with Robust Estimators Small versus Large Firms

Combined disclosure Social capital Human capital

Small fi rms Large fi rms Small fi rms Large fi rms Small fi rms Large fi rms

Share price volatilityFirm size 1 − ***−1.289 **−0.124 ***−1.304 ***−0.146 ***−1.439 ***−0.131Systematic risk 2 + 0.594 ***0.842 −0.464 ***0.819 −0.651 ***0.906Free fl oat 3 − 0.909 0.125 0.749 0.089 1.434 0.012Disclosure indicative 4 − 0.005 −0.005 −0.005 −0.003 −0.028 −0.023Disclosure qualitative 5 − 0.090 −0.010 0.088 −0.008 0.115 −0.040Disclosure quantitative 6 − **−0.663 0.012 *−0.563 −0.007 **−1.181 0.065Adjusted R2 44.6% 22.8% 41.8% 22.3% 42.1% 23.7%F statistic 5.00 (0.001) 10.8 (0.001) 5.65 (0.001) 13.2 (0.000) 4.53 (0.002) 10.8 (0.000)N: 143 33 110 33 110 33 110

Note: (Small fi rms if in the fi rst 25% percentile natural log of assets, large fi rms otherwise).



Table 8Ordinary Least Square Regression Model with Robust Estimators

Combined disclosure Social capital Human capital

Tobin’s QInverse of assets 1 + ***1.908 ***1.430 ***1.909Return on assets 2 + **2.552 ***2.309 ***2.977Disclosure indicative 3 + 0.004 −0.049 **0.556Disclosure qualitative 4 + −0.633 −0.328 −0.876Disclosure quantitative 5 + ***2.033 **1.543 **1.805Adjusted R2 33.2% 23.9% 43.1%F statistic 29.4 (0.000) 4.07 (0.001) 7.1 (0.000)N: 133

Note: *: p < 0.10; **: p < 0.05; ***: p < 0.01. One-tailed.

disclosure on share price volatility. Furthermore, it appears that fi rms take into account ultimate costs and benefi ts to shareholders when determining the attributes of their disclosure. Moreover, it appears that effi cient governance leads to more transparency in quantitative human capital disclosure while the extent of CEO stock options leads to less transparency in social capital disclo-sure, either soft or hard.

Contributions to Scholarship

We have shown how the market reacts to different levels of information precision and more specifi cally, to the differential impact of quantitative versus qualitative information disclosure. This fi nding supports an implicit assumption of recent studies regarding the differential weightings used to code corporate disclosure (e.g., Cho & Patten, 2007; Clarkson et al., 2008). Moreover, our results suggest the necessity of revisiting past empirical research on corporate social disclosure that relies on summary indicators combining all forms of information, thereby assuming that all disclosure is equally informative (e.g., Neu, Warsame, & Pedwell, 1998). Although quan-tifi cation is central in accounting, little is known about its effectiveness as persuasive information. In that sense, our study further extends the work of Clarkson et al. (2008) on the determinants of environmental disclosure to the context of social and human capital disclosures by assessing the value relevance of these disclosures.

Applied implications

Our fi ndings suggest that stock market participants perceive that only quantifi ed (hard) social capital disclo-sure reveals unknown information. The impact of social and human capital disclosure in reducing information

asymmetry seems to be more important for small fi rms than for larger fi rms. This may suggest that small fi rms more actively view the web as a cost-effective method to reach external constituencies such as investors. This is consistent with the fact that most traditional media (e.g., TV, newspapers, and magazines) carry only news from larger fi rms.

Limitations and Future Research Directions

The results reported here should be interpreted with caution for at least three reasons. Firstly, the paper focuses on HTML disclosure, which excludes hyper-linked documents in PDF. However, these documents (e.g., fi nancial statements, press releases, annual reports, sustainability reports, or proxy statements) are typically also published in paper form. Secondly, the use of a coding grid to quantify voluntary disclosure may be con-sidered subjective. However, our coding approach is con-sistent with recent studies (e.g. Pirchegger & Wagenhofer, 1999; Marston & Polei, 2004; Cho & Patten, 2007; Cormier et al., 2009). Thirdly, we focused on investors as the key benefi ciaries of enhanced social and human capital disclosure. To that end, the paper’s theoretical framework is inspired by information economics theory. However, fi rms may decide upon their disclosure strat-egy with a broader audience in mind, encompassing other stakeholders. Our results reinforce the view that future research on voluntary disclosure should take into account attributes of disclosure and not only its extent.

Note

1 Coding instructions as well as standardized coding work-sheets were prepared before hand. Each coder then applied



the following coding sequence: (1) independent identifi ca-tion of the occurrence of items relative to the different coding categories; (2) independent coding of the items according to quality level of content, and (3) timed recon-ciliation on a subset of company reports. The coders were trained in applying coding instructions and in using the coding worksheets. They were unaware of the hypotheses. Initial differences in identifying grid items accounted for an average 7% of the maximum number of items identifi ed. 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 fi rst 40 sample fi rms). A researcher reconciled coding disagreements exceeding 5% of the highest total score between the two coders. Smaller disa-greements were resolved by the two coders themselves.

References

Abbott, L.J., Park, Y., & Parker, S. (2000). The Effect of Audit Committee Activity and Independence on Corporate Fraud. Managerial Finance, 26(11), 55–67.

Aboody, D., & Kasznik, R. (2000). CEO Stock Option Awards and Voluntary Corporate Disclosure. Journal of Account-ing and Economics, 29, 73–100.

Adler, P., & Kwon, S.W. (2002). Social capital: prospects for a new concept. Academy of Management Review, 27(1), 17–40

Aerts, W., Cormier, D., & Magnan, M. (2007). The Association between Web-Based Corporate Performance Disclosure and Financial Analyst Behaviour under Different Gover-nance Regimes. Corporate Governance—An International Review, 15(6), 1301–1328.

Aerts, W., Cormier, D., & Magnan, M. (2008). Corporate Envi-ronmental Disclosure, Financial Markets and the Media: An International Perspective. Ecological Economics, 64, 643–659.

Al-Tuwaijri, S., Christensen, T.E., & Hughes II, K.E. (2004). The Relations among Environmental Disclosure, Environ-mental Performance, and Economic Performance: A Simultaneous Equations Approach. Accounting, Organi-zations and Society, 29, 447–471.

Anderson, U., Kadous, K., & Koonce, L. (2004). The Role of Incentives to Manage Earnings and Quantifi cation in Auditors’ Evaluations of Management-Provided Informa-tion. Auditing: A Journal of Practice & Theory 23(1), 11–27.

Atiase, R. (1985). Predisclosure Information, Firm Capitaliza-tion, and Security Price Behavior around Earnings Announcements. Journal of Accounting Research, 23 (Spring), 21–36.

Blair, M.M., & Kochan, T.A. (2000). The New Relationship: Human Capital in the American Corporation. Washing-ton, DC: Brookings Institution Press.

Botosan, C.A. (1997). Disclosure Level and the Cost of Equity Capital. The Accounting Review, 72(3), 323–350.

Botosan, C., & Plumlee, M. (2005). Assessing Alternative Proxies for the Expected Risk Premium. The Accounting Review, 80(1), 21–53.

Beasley, M.S. (1996). An Empirical Analysis of the Relation between the Board of Director Composition and Financial Statement Fraud. The Accounting Review, 71(4), 433–465.

Bédard, J., Chtourou, S., & Courteau, L. (2004). The Effect of Audit Committee Expertise, Independence and Activity on Aggressive Earnings Management. Auditing: A journal of Practice and Theory, 23(September), 13–35.

Burt, R.S. (1992). Structural Holes. Cambridge, M.A.: Harvard Business Press.

Chenhall, R., & Moers, F. (2007). The Issue of Endogeneity within Theory-Based, Quantitative Management Account-ing Research. European Accounting Review, 16(1), 173–195.

Chen, C.J.P., & Jaggi, B. (2000). Association between Indepen-dent Non-Executive Directors, Family Control and Finan-cial Disclosures in Hong Kong. Journal of Accounting and Public Policy, 19, 285–310.

Cheng, E., & Courtenay, S. (2005). Board Composition, Regu-latory Regime and disclosure. Illinois International Summer Accounting Conference, Japan.

Cheng, P.Y.K., Gopinath, S., & Krishnamurti, C. (2002). Does Frequent Trading Always Improve Liquidity? Working paper, ssrn.abstract=314384.

Cho, C.H., & Patten, D.M. (2007). The Role of Environmental Disclosures as Tools of Legitimacy: A Research Note. Accounting, Organizations and Society, 32(7/9), 639–647.

Clarkson, P., Kao, J.L., & Richardson, G.D. (1994). The Vol-untary Inclusion of Forecasts in the MD&A Section of Annual Reports. Contemporary Accounting Research, 11 (Fall), 423–450.

Clarkson, P., Li, Y., Richardson, G.D., & Vasvari, F.P. (2008). Revisiting the Relation between Environmental Per-formance and Environmental Disclosure: An Empirical Analysis. Accounting, Organizations & Society, 33, 303–327.

Cohen, D., & Prusak, L. (2001). In Good Company—How Social Capital Makes Organizations Work. Boston, MA: Harvard Business School Press.

Cormier, D., Ledoux, M.J., & Magnan, M. (2009). The Use of Web Sites as a Disclosure Platform for Corporate Perfor-mance. International Journal of Accounting Information Systems, 10, Forthcoming.

Cormier, D., & Magnan, M. (2003). Environmental Reporting Management: A European Perspective. Journal of Account-ing and Public Policy, 22, 43–62.

Dess, G.G., & Shaw, J.D. (2001). Voluntary Turnover, Social Capital, and Organizational Performance. Academy of Management Review, 26(3), 446–456.

Dye, R. (1985). Disclosure of Nonproprietary Informa-tion. Journal of Accounting Research, 23(Spring), 123–145.

Eng, L.L., & Mak, Y.T. (2003). Corporate Governance and Voluntary Disclosure. Journal of Accounting and Public Policy, 22, 325–345.

Fama, E.F. (1980). Agency Problems and the Theory of the Firm. Journal of Political Economy, 88(2), 288–307.



Fama, E.F., & Jensen, M.C. (1983). The Separation of Owner-ship and Control. The journal of Law and Economics, 26 (June), 301–325.

Fombrun, C., & Shanley, M. (1990). What’s in a Name? Repu-tation Building and Corporate Strategy. Academy of Man-agement Journal, 33, 233–258.

Francis, J., Khurana, I., & Pereira, R. (2005). Disclosure Incen-tives and Effects on Cost of Capital. The Accounting Review, 80(4), 1125–1162.

Frankel, R., McNichols, M., & Wilson, G.P. (1995). Discretion-ary Disclosure and External Financing. The Accounting Review, 70(1), 135–150.

Gebhardt, W., Lee, C., & Swaminathan, B. (2001). Toward an Implied Cost of Capital. Journal of Accounting Research, 39(1), 135–176.

Hail, L., & Leuz, C. (2006). International Differences in the Cost of Equity Capital: Do Legal Institutions and Securi-ties Regulation Matter. Journal of Accounting Research, 44(3), 485–531.

Hanlon, M., Rajgopal, S., & Shevlin, T. (2003). Are Executive Stock Options Associated with Future Earnings? Journal of Accounting and Economics, 36(1), 3–43.

Healy, P., Hutton, A.P., & Palepu, K.G. (1999). Stock Perfor-mance and Intermediation Changes Surrounding Sus-tained Increases in Disclosure. Contemporary Accounting Research, 16(3), 485–520.

Healy, P.M., & Palepu, K.G. (2001). Information Asymmetry, Corporate Disclosure, and Capital Markets: A Review of Empirical Disclosure Literature. Journal of Accounting and Economics, 31(1/3), 405–440.

Hill, C.W.L. (1990). Cooperation, Opportunism, and the Invis-ible Hand: Implications for Transaction Cost Theory. Academy of Management Review, 15, 500–513

Hirst, D., Koonce, L., & Simko, P. (1995). Investor Reactions to Financial Analysts’ Research Reports. Journal of Accounting Research, 33(2), 335–351.

Hitt, M.A., Lee, H.U., & Yucel, E. (2002). The Importance of Social Capital to the Management of Multinational Enterprises: Relational Networks Among Asian and Western Firms. Asia Pacifi c Journal of Management, 19, 353–372.

Hodge, F., Hopkins, P.E., & Pratt, J. (2006). Manage-ment Reporting Incentives and Classifi cation Credibility: The Effects of Reporting Discretion and Reputation. Accounting, Organizations and Society, 31(7), 623–634

Hope, O.K. (2003). Disclosure Practices, Enforcement of Accounting Standards and Analysts’ Forecasts Accuracy: An International Study. Journal of Accounting Research, 41(2), 273–272.

Hutton, A., Miller, G., & Skinner, D. (2003). The role of sup-plementary statements with earnings forecasts. Journal of Accounting Research, 41(5), 867–890.

Industry Canada (2008). Corporate Social Responsibility: The Importance of Stakeholder Engagement. Ottawa: Canada. (www.ic.gc.ca)

Ittner, C., & Larcker, D. (1998). Are Non-Financial Measures Leading Indicators of Financial Performance? An analysis

of Customer Satisfaction. Journal of Accounting Research, 36 (Supplement), 1–35.

Kaplan, R.S., & Norton, D.P. (1996). The Balanced Scorecard: Translating Strategy into Actions. MA: Harvard Business Scholl Press.

Kim, O., & Verrecchia, R. (1994). Market Liquidity and Volume around Earnings Announcements. Journal of Accounting and Economics, 17(1), 41–68.

Lim, T. (2001). Rationality and Analysts’ Forecast Bias. Journal of Finance, LVI (1), 369–385.

Lang, M., & Lundholm, R. (1993). Cross-Sectional Determi-nants of Analyst Ratings of Corporate Disclosures. Journal of Accounting Research, 31(2), 246–271.

Lang, M., & Lundholm, R. (1996). Corporate Disclosure Policy and Analyst Behavior. The Accounting Review, 71, 467– 492.

Leuz, C., & Verrecchia, R. (2000). The Economic Conse-quences of Increased Disclosure. Journal of Accounting Research, 38 (Supplement), 91–124.

Marston, C.L., & Polei, A. (2004). Corporate Reporting on the Internet by German Companies. International Journal of Accounting Information Systems, 5(3), 285–311.

McGuire, J., Sundgren, A., & Schneeweis, T. (1988). Corporate Social Responsibility and Firm Financial Performance. Academy of Management Journal, (December), 854–872.

Mercer, M. (2004). How Do Investors Assess the Credibility of Management Disclosures? Accounting Horizons, 18(3), 185–196.

Milgrom, P., & Roberts, J. (1992). Economics, Organizations, and Management. Englewood Cliffs, NJ: Prentice Hall.

Neu, D., Warsame, H., & Pedwell, K. (1998). Managing Public Impressions: Environmental Disclosures in Annual Reports. Accounting, Organizations and Society, 23(3), 265–282.

Nunnaly, J. (1978). Psychometric Theory. McGraw Hill, 2nd Edition, New York.

Orlitzky, M., & Benjamin, J.D. (2001). Corporate Social Per-formance and Firm Risk: A Meta-Analytical Review, Busi-ness & Society, 40(4), 369–396.

Pastoriza, D., Arino, M.A., & Ricart, J.E. (2008). Ethical Managerial Behaviour as an Antecedent of Organiza-tional Social Capital. Journal of Business Ethics, 78, 329–341.

Pennings, J.M., Lee, K., & Van Witteloostuijn, K. (1998). Human Capital, Social Capital and Firm Dissolution. Academy of Management Journal, 41(4), 425–440.

Pirchegger, B., & Wagenhofer, A. (1999). Financial Informa-tion on the Internet: A Survey of Homepages of Austrian Companies. The European Accounting Review, 8(2), 383–395.

Pyndick, R.S., & Rubinfeld, D.L. (1991). Econometric Models and Economic Forecasts. McGraw-Hill/Irwin.

Richardson, A., & Welker, M. (2001). Social Disclosure, Finan-cial Disclosure and the Cost of Equity Capital. Account-ing, Organizations and Society, 26(7), 597–616.

Robb, S.W.G., Single, L.E., & Zarzeski, M.T. (2001). Nonfi -nancial Disclosures across Anglo-American Countries.



Journal of International Accounting, Auditing and Taxa-tion, 10, 71–83.

Scott, T. (1994). Incentives and Disincentives for Financial Disclosure: Voluntary Disclosure of Defi ned Benefi t Pension Plan Information by French Firms. The Account-ing Review, 69(1), 26–43.

Sengupta, P. (1998). Corporate Disclosure Quality and the Cost of Debt. The Accounting Review, 73(4), 459–474.

Shane, S., & Cable, D. (2002). Networks Ties Reputation and the Financing of New Ventures. Management Science, 48(3), 364–381.

Skinner, D. (1994). Why Firms Voluntarily Disclose Bad News. Journal of Accounting Research, 32 (Spring), 38–60.

Standard & Poors (2002). Transparency and Disclosure: Over-view of Methodology and Study Results—United States. New York.

Stock Guide Publications (2005). Stock Guide, Division of Inovestor Inc. Montreal.

Verrecchia, R. (1983). Discretionary Disclosure. Journal of Accounting and Economics, 5, 179–194.

Verrecchia, R.E. (1990). Information Quality and Discretionary Disclosure. Journal of Accounting and Economics, 12, 365–380.

Waddock, S.A., & Graves, S.B. (1997). The Corporate Social Performance-Financial Performance Link. Strategic Man-agement Journal, 18, 303–319.

Welker, M. (1995). Disclosure Policy, Information Asymmetry and Liquidity in Equity Markets. Contemporary Account-ing Research, 11(3), 801–828.

Wiseman, J. (1982). An Evaluation of Environmental Disclo-sures Made in Corporate Annual Reports. Accounting, Organizations and Society, 7(4), 53–64.

Yhim, H.P., Karim, K.E., & Rutledge, R.W. (2003). The Asso-ciation between Disclosure Level and Information Quality: Voluntary Management Earnings Forecasts. Applied Financial Economics, 13(9), 677–692.

Xie, B., Davidson, W., & DaDalt, P. (2003). Earnings Manage-ment and Corporate Governance: The Role of the Board and the Audit Committee. Journal of Corporate Finance, 9(3), 295–316.

Xue, Y. (2007). Make or Buy New Technology: The Role of CEO Compensation Contract in a Firm’s Route to Innova-tion. Review of Accounting Studies, 12, 659–690.

Appendix

Disclosure Grid

Social capital disclosure Human capital disclosurePurchases of goods and services Hiring new employeesEmployment opportunities Qualifi cation expertiseJob creation TrainingEquity programs Description of job requirements 1, 2, 3Human capital development Total competenceRegional development Employee empowerment involvementGifts and sponsorships Capacity to suggest and to implement changesAccidents at work TeamworkHealth and safety programs Performance assessmentProduct-related-incidents Performance based compensationProducts in development and environment Earnings-based compensationProduct safety Carrier opportunitiesBusiness ethics AwardStrategic alliances Fringe benefi tsCommunity involvement Total motivation/work climateSocial activities Employees satisfaction, surveyOther Employee turnover

OtherTotal satisfaction

attributes of social and human capital disclosure and information asymmetry between managers and...

Documents