PROCESO DE ELABORACIÓN DEL QUESO CREMA

Scrutinised Contexts

This section mainly aims to further investigate whether there are significant differences in the influence of corporate governance related variables on the extent of compliance with mandatory IFRSs disclosure requirements between Egypt and Jordan. In order to investigate this issue, country dummy will be introduced and interacted with all test and control variables as follows:

Compliancej= a0+ a1 board independencej + a2 board leadershipj + a3 board sizej+ a4 government ownership ratioj + a5 management ownership ratioj + a6 private ownership ratioj + a7 public ownership ratioj + a8 Sizej + a9 Profitabilityj + a10 Gearingj + a11 Liquidityj + a12 type of business activityj + a13 type of audit firmj + a14 Country + a15 Country * board independencej + a16 Country * board leadershipj + a17 Country * board sizej+ a18 Country * government ownership ratioj + a19 Country * management ownership ratioj + a20 Country * private ownership ratioj + a21 Country * public ownership ratioj + a22 Country * Sizej + a23 Country * Profitabilityj + a24 Country * Gearingj + a25 Country * Liquidityj + a26 Country * type of business activityj + a27 Country * type of audit firmj + Ɛj

Where:

Compliancej = Total disclosure index (j=1,…, 150) a0= The intercept

a1 to a7 =Test variables a8 to a13 = Control variables a14=Country dummy

a15 to a21 = Country *Test variables a22 to a27 = Country *Control variables Ɛj= Error term

Table 6.18 demonstrates the Stepwise regression results for total disclosure index41.

Table 6.18: Significant Differences between Scrutinised Contexts in the Influence of Independent Variables on

the Levels of Compliance with IFRSs Disclosure Requirements

**, *** Significant at the 0.05 and 0.01 Levels respectively

As demonstrated in Table 6.17 and Appendix 9 the VIF and Tolerance values indicate the non- existence of the problem of multicollinearity (i.e., all VIF values are less than 10 and all Tolerance

41

Coeffecient estimates and significance level for independent variables that are excluded from the regression model are presented in Appendix 9.

Model summary

R R2 Adjusted R2 Std. Error F Sig.

.449 .202 .191 .877 18.565 .000***

Coefficients Collinearity Statistics

Predictor Variable Unstandardised Coefficients Standardised Coefficients t Sig. Tolerance VIF B Std. Error Beta (Constant) .242 .085 2.842 .005*** Country*Auditor -.844 .161 -.393 -5.252 .000*** .970 1.031 Firm Size .160 .075 .160 2.139 .034** .970 1.031

values are above .2). The adjusted R2 = .191, implying that 19.1% of the variation in the total disclosure index (overall compliance with mandatory IFRSs disclosure requirements) for the entire sample of companies listed on scrutinised MENA stock exchanges, is explained by the regression model. Findings indicate a significantly higher influence of auditing by big 4 audit firms on the levels of compliance with mandatory IFRSs disclosure requirements for companies listed on the ASE than for those listed on the EGX. The interactive variable Country*Type of auditor shows a significant coefficient at .01 level. This result is reasonable as big 4 audit firms' staff are more qualified and have better experience with IFRSs, and extensive training on any updates of these standards than auditors in local audit firms who may have limited experience and knowledge particularly in a country like Jordan that is suffering from a shortage of qualified accountants and auditors. Additionally, big 4 audit firms exert more pressure on their clients to improve compliance with IFRSs. This result supports the findings of many earlier studies (e.g., Singhvi & Desai, 1971; Ahmed & Nicholls, 1994; Raffournier, 1995; Wallace & Naser, 1995; Inchausti, 1997; Patton & Zelenka, 1997; Abd-Elsalam, 1999; Glaum & Street 2003; Samaha & Stapleton, 2009). In addition, it supports those of Suwaidan (1997), Naser et al. (2002), Omar (2007) and Al-Akra et al. (2010a) in the Jordanian context.

The coefficient firm size is positive and significant at .05 level for the entire sample. This implies that large companies comply better with IFRSs disclosure requirements (the coefficient estimate is significant at .05). This confirms the findings of the majority of prior researchers (e.g., Inchausti, 1997; Depoers, 2000; Haniffa & Cooke, 2002; Eng & Mak, 2003; Akhtaruddin, 2005; Naser et al., 2006). It also supports the findings of Suwaidan (1997), Naser (1998), Abd-Elsalam (1999), Naser et al. (2002), Omar (2007), Samaha and Stapleton (2009) and Al-Akra et al. (2010b) in the scrutinised contexts. Referring to political cost theory, many researchers argue that large companies are more likely to provide more disclosures and to comply with mandatory disclosure requirements as they may be more subject to public scrutiny or more sensitive to criticism for falling short of disclosure requirements compared to small firms (e.g., Singhvi, 1968; Singhvi & Desai, 1971; Busby, 1975; Firth, 1979; Abd-Elsalam, 1999; Al-Htaybat, 2005). Referring to agency theory, many researchers propose that large companies are more likely to disclose more information because of large numbers of shareholders and the associated pressures (e.g., Hossain et al., 1995; Meek et al., 1995; Al-Mulhem, 1997; Abd-Elsalam, 1999; Ali et al., 2004; Al-Htaybat, 2005; Omar, 2007). Also, using capital need theory, many researchers claim that larger companies are more likely to disclose more information in order to raise funds at lower costs (e.g., Cooke, 1991; Meek et al., 1995; Abd-Elsalam, 1999; Ali et al., 2004; Al-Htaybat, 2005; Omar, 2007). Furthermore, using cost-benefit analysis, many researchers argue that large firms are more likely to disclose more information because for them, the costs of non-compliance are higher than the costs of compliance. Also, large companies are more likely to disclose more information because of lower

costs associated with collecting and publishing information and limited impact on the competitive position compared to small companies (e.g, Singhvi, 1968; Singhvi & Desai, 1971; Busby, 1975; Firth, 1979; Ahmed & Nicholls, 1994; Abd-Elsalam, 1999; Ali et al., 2004; Al-Htaybat, 2005). Additionally, based on the notions of the institutional isomorphism theory, large companies are more likely to comply with IFRSs to gain legitimacy and respect, hence attract foreign investors.

With respect to the association between board independence and levels of compliance with IFRSs disclosure requirements, regression results support the non-existence of a statistically significant relationship between board independence and the overall level of compliance with mandatory IFRSs disclosure requirements for the entire sample. Hence, these findings support the second research hypothesis (H2: there is no significant statistical relationship between board independence and the extent of compliance with IFRSs disclosure requirements). This agrees with the findings of Haniffa (1999), Ho and Wong (2001), Haniffa and Cooke (2002), Ghazali and Weetman (2006). In addition, this result supports those of Al-Akra et al. (2010a,b) in the Jordanian context, however, it does not support those of Samaha and Dahawy (2010; 2011) which report a significant positive relationship between voluntary disclosure levels and board independence in the Egyptian context. This supports the proposition that, most outside directors in the scrutinised listed companies, lack material independence as generally, they are appointed to the board because of their close relationship with executive board members, the Chair or controlling shareholders. They may also lack experience or may have insufficient financial incentive to actively monitor management and protect the interests of minority shareholders. This lends weight to the notions of institutional isomorphism theory (board members in the scrutinised listed companies, do not contribute to improving the BOD’s monitoring function even when they meet the independence criterion, being appointed simply to signal that such companies follow corporate governance best practices, and hence, gain legitimacy and respect). Furthermore, the predictions of financial economics theories relevant to this study (weak monitoring reduces monitoring costs, hence will not stimulate management to improve compliance, as well as the weak enforcement of compliance with IFRSs results in low non-compliance costs) and Gray’s (1988) accounting sub-cultural model (acceptance of secrecy as a dominant culture in the scrutinised contexts and absence of awareness regarding the importance of transparency support management’s selective disclosure to avoid competition and protect company reputation, even though the lack of disclosure is in breach of the mandatory requirements). Moreover, the interviewees' responses discussed in Chaprter Seven also support this proposition.

With respect to the impact of board leadership, regression analysis reveals that holding the CEO and Chair positions by two different persons does not improve the monitoring function.

Consequently, H3: there are no statistically significant differences in the levels of compliance with IFRSs disclosure requirements between companies that separate the positions of the CEO and Chair and those that do not, is accepted. This supports the findings of Arcay and Vazquez (2005), Cheng and Courtenay (2006), Ghazali and Weetman (2006) and Ezat and El-Masry (2008). This finding may be attributed to the lack of material independence of the Chair in the scrutinised listed companies when the CEO and the Chair positions are separated. This in turn lends support to the institutional isomorphism which suggests that, separating the CEO and Chair positions has no influence on board leadership independence, as long as there is no awareness regarding the importance of separating the positions in improving the monitoring of management behaviour and hence the quality of financial reporting within the business firm. Consequently, no significant impact on levels of compliance with IFRSs, is expected when the two positions are separated, and decoupling is thus expected to continue due to the existence of cultural barriers to understanding the logic behind the separation of the two positions as recommended under the Anglo-American model of corporate governance. In addition, companies may fall in line with the separation recommendations purely to gain respect and legitimacy. Also, Gray’s (1988) accounting sub- cultural model, the notions of agency theory and cost benefit-analysis would argue that given the dominant secretive culture accompanied with lack of material independence, weak monitoring and lack of strict enforcement of compliance, non-compliance costs will continue to be less than compliance costs. Consequently, the separation between the CEO and Chair positions may not result in better compliance with IFRSs disclosure requirements. This viewpoint is confirmed by the interviewees' responses reported in Chapter Seven.

On the other hand, regression results reveal the non-existence of association between board size and compliance with overall IFRSs disclosure requirements for the entire sample. Hence, H4: there is no significant statistical relationship between board size and the extent of compliance with IFRSs disclosure requirements, is accepted. This agrees with the findings of Lakhal (2003), Arcay and Vazquez (2005) and Cheng and Courtenay (2006). However, this result does not support that of Al- Akra et al. (2010a) who investigated this issue in the Jordanian context and reported a significant positive relationship. This finding can mainly be explained by the lack of independence of board members and the lack of awareness among them with respect to their role in monitoring management behaviour and protecting the interests of shareholders.This supports the notions of institutional isomorphism theory (board members in the scrutinised contexts, do not contribute to improving the BOD’s monitoring function even when they meet the independence criterion, being appointed simply to signal that companies follow corporate governance best practices, and hence, gain respect). Furthermore, weak enforcement and monitoring by the regulatory bodies, reduced agency costs, dominance of secretive culture and lack of understanding by board members

regarding the benefits of transparency, will not stimulate management to improve levels of compliance with IFRS.

With respect to the association between government ownership ratio and levels of compliance with IFRSs disclosure requirements, the regression results reveal the non existence of association. Hence, hypothesis H5a (there is no significant statistical relationship between the government ownership ratio and the extent of compliance with IFRSs disclosure requirements), is accepted. This result supports the findings of Ghazali and Weetman (2006) as well as the findings of Samaha and Dahawy (2010; 2011) on the level of the Egyptian context. In addition, it supports that of Naser et al. (2002), Al-Akra et al. (2010a) in the Jordanian context.This may be explained by the government’s ability to access all company information. Agency theory suggests this reduces the monitoring costs, and hence reduces management incentives to improve disclosure. Simultaneously, the notions of Gray’s (1988) accounting sub-cultural model, cost-benefit analysis, and institutional isomorphism are supportd. Given the preference to secrecy the government may not encourage full transparency. This may be due to government’s intention to sell its shares in the company at a good price as part of the privatisation programme. Additionally, the lack of awareness and the absence of incentives for members of the public who are implicit owners of government shares discourage direct monitoring of the management (generally government officials) in companies with dominant government ownership. This contributes to the decoupling problem as companies declare their compliance with IFRSs when in reality they are not complying, simply to gain respect and legitimacy. Further support for this result is evident in the interviewees’ responses presented in Chapter Seven.

With respect to the association between management ownership ratio and levels of compliance with the overall IFRSs disclosure requirements, regression results do not support H5b (there is a significant negative statistical relationship between the management ownership ratio and the extent of compliance with IFRSs disclosure requirements). Although, this result does not agree with the findings of the majority of prior studies investigating the association between management ownership and levels of disclosure which support a negative relationship (e.g., Eng & Mak, 2003; Arcay & Vazquez, 2005; Ghazali & Weetman, 2006; Abdelsalam & El-Masry, 2008; Samaha & Dahawy, 2010), it supports that of Samaha and Dahawy (2011) in the Egyptian context. This may be explained by the lack of separation between ownership and control, reduced agency costs, and predictions based on Gray’s (1988) accounting sub-cultural model, cost-benefit analysis, and institutional isomorphism that the secretive culture, lack of management awareness concerning the importance of transparency and compliance, absence of monitoring by non-executive board members or stock exchange regulators, and the absence of pressure from minority shareholders, encourage management to keep disclosure levels at a minimum as long as non-compliance costs are

less than compliance costs. This in turn contributes to the problem of decoupling. Further support for this result is evident in the interviewees’ responses presented in Chapter Seven.

With respect to the association between private ownership ratio and levels of compliance with IFRSs disclosure requirements, regression results support H5c (there is no significant statistical relationship between the private ownership ratio and the extent of compliance with IFRSs disclosure requirements). This supports the findings of Depeors (2000); and those of Suwaidan (1997), Omar (2007) and Al-Akra et al. (2010a,b) in the Jordanaian context, however, it does not support those of Samaha and Dahawy (2010; 2011) with respect to voluntary disclosure practices in Egypt. This can be explained by ease of access to all company information by private investors who are in most cases actively involved in company management either as executives or as directors. Moreover, based on Gray’s (1988) accounting sub-cultural model, cost-benefit analysis, and institutional isomorphism, it can be stated that the secretive culture, and lack of private investor awareness of the importance of transparency will not increase pressures by private investors on management to improve compliance with IFRSs. Additionally, absence of monitoring from board members or stock exchange regulators, and the absence of pressure from minority shareholders will reduce non compliance costs; hence will not improve compliance levels with mandatory IFRSs. Consequently, this contributes to the problem of decoupling. Further support for this result is evident in the interviewees’ responses presented in Chapter Seven.

With respect to the association between public ownership ratio and levels of compliance with IFRSs disclosure requirements, regression results support H5d (there is no significant statistical relationship between public ownership ratio and the extent of compliance with IFRSs disclosure requirements). This result supports that of Hossain et al. (1994). Furthermore, it supports that of Naser et al. (2002) and Al-Akra et al. (2010b) in the Jordanian context. However, it does not support the results of the majority of prior studies which report the existence of a positive association between public ownership and disclosure level (e.g., Haniffa, 1999; Haniffa & Cooke, 2002; Arcay & Vazquez, 2005). In addition, it does not support the findings of Ezat and El-Masry (2008) in the Egyptian context and those of Al-Htaybat (2005) in the Jordanian context.This result may be attributed to the lack of demand for more disclosure by public investors in the scrutinised contexts. In this regard, Naser et al. (2002) and Al-Akra et al. (2010a) argue that, dominance of public ownership in the Jordanian context results in lower monitoring capacity as the majority of individual investors in Jordan are small unsophisticated investors, and their investment decions in most of the cases are speculative and uninformed (Naser et al., 2002; Al-Akra et al., 2010a,b). In a similar vein, Abdelsalam and Weetman (2007) describe public shareholders in Egypt as small investors who cannot form pressure groups like those in developed markets. Reduced agency costs

due to the lack of demand for more disclosure by public investors, accounts for this. Furthermore, the lack of listed companies’ management and BOD awareness regarding the importance of compliance with IFRSs and of following corporate governance best practices to enhance transparency, the weak enforcement of laws and regulations, and the absence of materially independent board members with primary responsibility for protecting publicshareholders' interests, cause non-compliance costs to be less than compliance costs. The fact that public shareholders in scrutinised stock exchanges do not exercise their rights, adds to this situation, thereby management is not stimulated to improve compliance with IFRSs, and the problem of decoupling escalates. Further support for this result is evident in the interviewees’ responses presented in Chapter Seven.

With respect to control variables, regression results for the entire sample reveal that, no association exists between firm profitability and its level of compliance with IFRS disclosure requirements in scrutinised contexts, thereby supporting the findings of some previous studies (e.g., Malone et al., 1993; Al-Mulhem, 1997; Inchausti, 1997; Tower et al., 1999; Ho & Wong, 2001; Eng & Mak, 2003; Barako et al., 2006). Additionally, this supports the findings of Abd-Elsalam (1999), Ezat and El-Masry (2008), Al-Akra et al. (2010a,b) and Samaha and Dahawy (2010; 2011) in scrutinised contexts. Regression results also confirm the non-existence of association between firm gearing and its level of compliance with IFRSs disclosure requirements, thereby supporting the outcomes of some prior research (e.g., Hossain et al., 1994; Raffournier, 1995; Wallace & Naser, 1995; Inchausti, 1997; Patton & Zelenka 1997; Craig & Diga 1998; Dumontier & Raffournier 1998; Tower et al., 1999; Ho & Wong, 2001; Haniffa & Cooke, 2002; Aksu & Kosedag 2006). In addition, they support those of Abd-Elsalam (1999), Al-Htaybat (2005), Ezat and El-Masry (2008), Al-Akra et al. (2010b) and Samaha and Dahawy (2010; 2011) in the scrutinised contexts. Similarly, regression results reveal the non-existence of association between company liquidity and its level of compliance with IFRS disclosure requirements in scrutinised contexts. This supports the findings of Alsaeed (2005), and Barako et al. (2006). Additionally, it supports the findings of Abd-Elsalam (1999) in the the Egyptian context. It also supports those of Al-Htaybat (2005) in the Jordanain context. Hence, compliance behaviour with mandatory IFRSs disclosure requirements by companies listed on the scrutinised stock exchanges is not influenced by company liquidity. Additionally, the regression results show that, the type of business activity does not affect the level of compliance with mandatory IFRSs disclosure requirements in the scrutinised contexts. This finding supports those of some prior studies (e.g., Inchausti, 1997; Street & Gray, 2002). In addition, it supports the findings of Ismail et al. (2010) and Samaha and Dahawy (2011) in the Egyptian context, and those of Naser (1998) and Naser at al., (2002) in the Jordanian context. The