BARRIERS OF GOOD CORPORATE GOVERNANCE PRACTICES: EVIDENCE FROM EMERGING ECONOMY

Corporate governance (CG) is often split among rule and principle-based methods to regulation in distinctive institutional contexts. CG practices are divergent among countries due to differences in their institutional environments. Therefore, the purpose of this study is to identify the barriers of CG practices and compliance in Pakistan. The primary data is collected from 105 Pakistan Stock Exchange (PSX) listed firms through survey questionnaire. The study employed the exploratory factor analysis (EFA) to identify the barriers of good CG practices and compliance in Pakistan. Using exploratory factor analysis (EFA), this study identified five major barriers, i.e. firm level barriers (lack of auditor independence, board ineffectiveness, lack of shareholders’ awareness), external barriers (political and governmental interference in business activities, weak legal control and enforcement, high levels of corruption), social barriers (strong social ties among different stakeholders, interpersonal connections among boards of directors (BoDs), education and training barriers (lack of professional education and training among stakeholders) and legal barriers (fewer voting rights) which restrain good CG practices in Pakistan. The study contributes to existing CG literature and highlights the barriers which hinder the good CG practices and compliance, especially in the context of Pakistan. The findings are applicable to other emerging markets especially in Asian countries due to similar socio-economic environment.


BACKGROUND OF THE STUDY
Recently, the corporate governance (CG) has gained considerable attention from academics and regulators due to mega corporate scandals.These incidents reflect the lack of CG practices both at firm and country level.Researchers argued that CG practices are divergent among countries due to differences in their institutional environments [1] and many barriers hinder the CG practices.Consequently, this has led to a growing appreciation of the institutional effects on CG in developed countries (2-6), however, a comparative gap still exists in the CG literature for developing countries that are usually characterized by weak institutional environment [7].
Many developing countries have introduced the CG practices, adopted from developed countries, but their socio-economic environment is quite different from those developed countries [8].Hence, CG compliance is not up to mark in those developing countries.
Similarly, in Pakistan, the Security and Exchange Commission of Pakistan (SECP) introduced the corporate governance code (CCG) in 2002 and a reform in 2012.However, the researchers found that the CG compliance is still not up to the mark among PSX listed firms [9] and tick box practice is quite common (10,11).On the other side, researchers also highlighted the importance of institutional environment in the effectiveness of CG practices [3,7,12,13].
Researchers also argued that prevailing culture in emerging markets contributes to the weak CG practices in those countries [14] and suggests the adoption of cultural-cognitive system which may improve CG practices by harmonizing the key elements of normative and regulatory systems [15].It is also found that firms must understand and negotiate to different environmental influences, including politics and culture, for its survival [16].DiMaggio and Powell [17] argued that firms are similar to each other in different areas and this concept is called isomorphism.Firms can imitate the activities and operations of another firm in a certain environment to become similar [18].The institutional isomorphism can be manifested in three forms (i.e.Mimetic, Normative and Coercive).In mimetic isomorphism, a firm deliberately tries to imitate the practices of established competitors in highly unpredictable environment for survival.While in normative isomorphism, there is no deliberate attempt of a firms to imitate its established competitors.Normative isomorphism emerges due to engagement of operatives and managers from its competitors.Coercive isomorphism is demand for change due to formal and informal pressure from societal culture and other organizations upon which firms depend [17].This highlights the importance of institutional environment to CG.Similarly, Knack and Keefer [19] utilized the robustness of institutional environments to expound countries into developing and developed.Consequently, the developed countries should have relatively strong corporate governance system as compared to developing countries due to having sound institutional framework, however, the classic cases, such as Enron, highlights concern regarding corporate governance in developed countries [20].It implies that though the institutional environment is critical to corporate governance discourse, key players in a business could equally affect the emergence of (negative) isomorphic tendencies in a business environment.
Since 1980s, the agency paradigm of corporate governance has been dominated in the existing literature [21,22], however, a number of recent CG studies emphasized on more holistic view and accounted the organizations' nexus with society and different stakeholders [23].Aguilera and Jackson [3] argued that the agency theory is an "under-socialized" approach which is impervious to how institutions share the interests and identities among actors in CG system.In addition, agency theory only focuses on managers and shareholders.Though different countries offer distinctive level of investor protection and therefore, effect the agency costs [24], the role of institutions is very restricted from agency perspective.Sociological strands of institutional theory offers an alternative justification for firm behaviour in terms of "understandings that organizational actors share, independent of their interests" [17].Institutional theory does not make projections on based on self-interested actors with bounded rationality but ascertains the normative, regulative and cognitive mechanisms that shape the interests and identities broad range of stakeholders.Aoki et. al.,[25]

METHODOLOGY Population and Sample Size
The population of study consists of all the 579 firms listed on the Pakistan Stock Exchange (PSX) and the sample should be drawn carefully to represent the whole population.However, researchers also documented some determinants such as research objectives, time and cost, proposed analysis and size of population that may affect decisions regarding selection of the sample size [38,39].In addition, Kothari [40] suggested that the sample size should not be too large nor too small.Consequently, this study used a purposive sampling technique to recruit the sample.This technique has been widely used in the existing studies on CG in different countries [41][42][43][44].This is a type of non-probability sampling technique in which the sample is selected in view of the purpose and defined criteria [45].First, this study excludes the financial companies from the sample due to their different CG structure.Second, the survey was conducted from different respondents including managers, accountants, auditors or other members of organizations who were involved in the preparation of CG reports.Based on the above criteria, the questionnaire was distributed to 350 respondents, however, only 120 questionnaires were received.Out of 120 filled questionnaires, 15 questionnaires were incomplete, hence, making a final sample of 105 respondents.
The study used the survey questionnaire to collect the primary data from respondents.Collis and Hussey [46] documented that a questionnaire is a list of carefully chosen structured questions that are executed after considerable testing to elicit responses from respondents.The questionnaire needs to be in simple and concise language to ensure the respondents understand the meaning of questions in the same way.Moreover, questions need to be specific so that  As seen in Table 1, the Cronbach Alphas for barriers of good CG practices is above 0.70.The coefficient of Cronbach Alpha ranges between zero to one and above 0.7 are considered as highly reliable (39).The study also employed exploratory factor analysis (EFA) to identify the main barriers of good CG practices and compliance in Pakistan.

Results
The results of the demographic information of respondents are presented in Table 2.There was a total of 105 respondents.The result reveals that most respondents (39%) were aged between 31 and 40 years while only 3.8% respondents were above 60 years of age.Regarding position, the results reveal that most of the respondents (42.9%) held the position of manager while 33.3% were in senior manager positions.There were only 8.6% respondents in the position of auditor while 15.2% of respondents were in the position of accountant.Coupling the demographic information on age and position is a way of providing reliable data for steady analysis.Note: Author's own design and calculated based on the information (survey data) The results of the demographic information also reveal that 48.6% of respondents have a master's degree or equivalent qualification, while only 2.9% of respondents have other qualifications.It is highlighted that only 9.5% of respondents have PhDs or equivalent qualifications while Bachelors' and Diploma holders were 31.4 % and 7.6% respectively.
The results also reveal that most of the respondents (34.3%) have specialization in accounting while only 1.9% respondents have other specializations.There were 10.5 % respondents who have specialization in management while both finance and economic specialization holders are each 26.7%.Regarding experience, results reveal that most of the respondents (41%) have 11 to 15 years of experience while 40% respondents have 6 to 10 years of experience.The 18.1% respondents who were in their early career have 1 to 5 years of experience, while only one respondent has 16 to 20 years of experience.All in all, the respondents are mature experienced managers.

Barriers to Good CG Practices in Pakistan
This section presents the results of exploratory factor analysis (EFA) regarding barriers to good CG practices and compliance in Pakistan.The survey includes 17 items that were measured on the five-point Likert scale.The results of Kaiser-Meyer-Olkin (KMO) and Bartlett's tests are presented in Table 3. KMO measure the sampling adequacy and should be greater than 0.5 for performing a satisfactory factor analysis.Kaiser [52] provided guidelines for interpreting these values[53]2 .Table 3 reveals that the KMO has a value of 0.702 which shows the adequacy of the sample for EFA.In addition to KMO, the Bartlett's test of sphericity that estimates which inter-correlation matrix produced is an identity matrix.Generally, the value of P<0.05 on Bartlett's test indicates that the inter-correlation matrix is not an identity matrix and factor analysis can be performed.
In It is also noted that the number of factors retained varies across studies and different criteria have been used to make a decision.The following criteria (i.e.eigenvalue, scree test and variance explained) have been used in this study to make a decision about factor retention.Kaiser [59] suggested that only those factors are retained for interpretation that have eigenvalues greater than 1.0.Eigenvalues represents the explained variable by a given factor.
The benchmark of establishing eigenvalue 1.0 seems arbitrary, however, researchers documented that factors that have eigenvalues greater than one should be retained because these represent those factors which contribute to a higher percentage of communal variance than average [60].
In addition to eigenvalue, there is an alternative approach called the Scree test to determine factors' retention and involves developing a scree plot of extracted factors against the magnitude of their eigenvalues [61,62].In this approach, the researcher needs to identify an    Regarding the first factor (firm level barriers), the study finds that from the CG implementation viewpoint, auditors' independence is compromised in Pakistan.In addition, the BoDs have failed to be effective and shareholders' awareness acts as a barrier to implementation of good corporate governance practices in Pakistan.These findings are well supported by Capaul [71] who argued that abortive auditing is a key problem in CG enforcement in most transition and developing economies because these economies have failed to implement their rules and regulations.Auditors' independence is compromised, and they have been giving categorical opinions, verifying that the audited accounts represent true picture despite the presence of many defects.
Regarding the second factor (external barriers), the study finds political and governmental influence in firms and weak legal control and enforcement of regulatory bodies.In addition, the study also finds that the level of corruption is another major factor that hinders good

Conclusions and Recommendations
The study finds that there are different barriers such as firm level barriers (lack of auditor independence, board ineffectiveness, lack of shareholders awareness), external barriers Appointments are made on the basis of social ties and personal relationships.Hence, it is respondents do not give several answers.The questionnaire was formulated from multiple sources to ensure validity [47-49].The questionnaire consisted of two parts including demographic information to increase participants' confidence [50].The first part comprised of the demographic information of the respondents.The second part comprised the barriers of good CG practices and compliance in Pakistan, which is measured through a five-point Likert scale ranging from (i) strongly disagree to (v) strongly agree.The part two (barriers) consists of seventeen items; (i) Lack of Auditors' Independence, (ii) Board Ineffectiveness, (iii) Institutional Culture of Pakistan, (iv) Political and Governmental Interference in Business Activities, (v) Weak Legal Control and Enforcement, (vi) Lack of Shareholders' Awareness, (vii) Lack of Resources for CG Compliance, (viii) Lack of Shareholders' Rights Protection especially Minority Shareholders, (ix) Lack of Protection for Whistle Blowers, (x) Lack of Professional Education and Training among Stakeholders, (xi) Fewer Voting Rights, (xii) Low AGM Participation, (xiii) High Level of Corruption, (xiv) Nepotism or Kinship Culture, (xv) Wobbly/unstable Economy of Pakistan, (xvi) Strong Social Ties among Different Stakeholders, and (xvii) Interpersonal Connections among BoDs.The researcher hired two research assistants to distribute the questionnaire to the respondents to save time and cost 1 .The distribution and collection of questionnaires took about two months.The instruments were pre-tested to ensure the content and face validity by analysing consistency and interpretation.For this purpose, the questionnaire was sent to experts in the field of CG to eliminate ambiguity and inadequacy.Simple words and language were used to ensure validity.Moreover, redundant and complicated terminologies were eliminated.The items were tested for their reliability through Cronbach Alpha with the help of Statistical Package for Social Sciences (SPSS) 24.The Cronbach alpha indicates how well items in a set are positively correlated to each other and Cronbach alpha is determined for the items.The results of Cronbach Alpha are presented below ( elbow or break where larger eigenvalues end in steep slope rambling off of smaller eigenvalues begins.Cattell[62]  suggested that only left side factors of the elbow are retained while right side factors are dropped.Another common method about making a decision regarding factor retention is examining the cumulative variance accounted for by retained factors.Various sources recommended numerous levels from 50% onwards and there is no exact percentage of total variance explained.However, most statisticians and scholars recommended factors that are required to obtain a variance of 75% to 90% [see63, 64, 65].Based on the above criteria (i.e.Eigenvalue, Scree test and Total variance), this study only retained five factors that have eigenvalues of greater than one.The Scree test (Fig. 1) also reveals the elbow after five factors.In addition, Table 4 also reveals that these five factors explain the cumulative variance of 80.29% which is recommend by other researchers [see 63].In Table4, there are three main components, i.e. initial eigenvalues, extraction sums of squared loadings and rotation sums of squared loadings.The first factor has eigenvalues of 5.861 and these explain 34.476% of variance explained.Similarly, the second factor has an eigenvalue of 2.922 and explains 17.187% of variance, making a cumulative variance of 51.663%.The third factor has an eigenvalue of 2.279 and explains 13.408% of variance, making a cumulative variance of 65.072%.The fourth factor has an eigenvalue of 1.406 and explains 8.273% of variance and cumulative variances reach to 73.344%.Finally, the fifth factor has an eigenvalue of 1.182 and explains 6.954% of variance and cumulative variance reaches to 80.298% which is within the recommended range by the researchers [see 63].

Figure 1
Figure 1 Scree plot of EFA for Barriers to Good Corporate Governance practices Source: Done by author based on the information (survey data) Note: Done by author based on the information (survey data)In addition, an elbow can be seen after factor five.As suggested byCattell [62], only these five factors are retained in this study that were on left side of the elbow and the right side factors were dropped.It is often difficult to interpret factors that are initially extracted and retained.Consequently,Dimitrov [66]  recommended that researchers need to rotate these factors to more suitable positions to create the simplest possible factor structure, maximize high loading and minimize low loadings.The idea of rotating factors stems from the work of researchers such as Thurstone[67]  and Cattell[62]  who saw it as a way of simplifying factor structures so they could be more reliably interpreted.In SPSS, researchers have to choose either orthogonal or oblique rotation strategies which have quite different fundamental assumptions but have same goal (i.e.seeking simple structure)[63, 65].In orthogonal rotation, it is assumed that factors are independent of one another, consequently, they are kept in a fixed position and it is expected that newly rotated factors are uncorrelated.Varimax, quartimax and equamax are three common orthogonal rotation algorithms and varimax is the most widely used and easy to interpret [66].This study selected varimax from the rotation menu and chose to suppress factors, having a coefficient score of less than 0.50 due to small sample size 3 .Table5presents the rotated results of principal axial factoring for barriers to good CG practices and compliance in Pakistan.Factor loadings were considered in evaluating the factors retention that represent barriers of good CG practices and compliance in Pakistan and only those factors were retained which had a minimum factor loading of 0.7 that is considered excellent [see 53].Table 5 reveals that 17 items that were included into EFA were extracted and loaded into 5 factors.The factor 1 (three items) comprised lack of auditors' independence, board ineffectiveness and lack of shareholders' awareness.The factor 2 (three items) comprised political and governmental interference in business activities, weak legal control and enforcement and high levels of corruption.The factor 3 (two items) comprised strong social ties among different stakeholders and interpersonal connection among BoDs.The factor 4 (one item) comprised lack of professional education and training among stakeholders.At the end, factor 5 (one item) comprised fewer voting rights.Based on item loading and shared characteristics on each factor, the researcher assigned factor labels.The factor 1 is labelled as firm level barriers, factor 2 is 3 Due to the small sample, this study supressed the small coefficient of absolute value of 0.5.(see 68) BoDs), education and training barriers (lack of professional education and training among stakeholders) and legal barriers (fewer voting rights).These five barriers are the most important barriers which are affecting the implementation of good CG practices in Pakistan.The results are well supported from the previous studies [69, 70].
corporate governance practices in Pakistan.Researchers argued that the political system of a country can has implications for corporate governance[72].The findings of the study reveal that weak legal control and enforcement are barriers to implement good CG practices in Pakistan, while researchers documented that enforcement is vital for providing good CG system and an effective business environment in developing countries like Pakistan[73].Similarly,Wilson [74]  documented that firms can be estranged from the corruption that prevails in the society [75] if they are operating in a weakened corporate governance environment like Pakistan.Regarding the third factor (social barriers), the study finds that strong social ties and also interpersonal connections among BoDs, also hinder good corporate governance practices in Pakistan.Similarly, Haniffa and Cooke [76] found that social factors such as culture affect the CG practices in emerging countries.Regarding the fourth factor (education and training barrier), the study finds that stakeholders lack professional education and training.In a similar vein, Okpara [77] documented that lack of required education and training is a barrier which hampers the development and implementation of corporate governance practices in Pakistan.Regarding the fifth factor (legal barriers), this study finds that shareholders have fewer voting rights, consequently, this acts as a barrier in the implementation of CG practices in Pakistan.Due to limited voting rights, the protection of shareholders' rights is also absent in Pakistan.In a similar vein, Okpara [77] documented that shareholders' rights are very crucial and vary from country to country.In addition, Jiraporn and Davidson [78] argued that shareholders' rights are an important part of corporate governance and play a pivotal role in controlling the behaviour of BoDs.Researchers argued that there is a need to provide effective protection in law to disgruntled minority shareholders [79].

(
political and governmental interference in business activities, weak legal control and enforcement, high level of corruption), social barriers (strong social ties among different stakeholders, interpersonal connections among BoDs), education and training barriers (lack of professional education and training among stakeholders) and legal barriers (fewer voting rights) that restrain the CG practice and compliance in Pakistan.Government of Pakistan(GOP) and regulatory bodies (SECP and Stock Exchange) need to take appropriate measures to control these barriers in upcoming CG reforms and codes.The findings of study reveal that auditing process in not effective in Pakistan, hence, government needs to make strict criterion regarding appointment of internal and external auditors.In Pakistan, CCG 2012 required at least one independent director while it is increased to two independent directors in new CCG 2017.The problem is not about numbers, it is about true and fair compliance of CG code.It is proposed that SECP and Pakistan Institute of Corporate Governance (PICG) need to set a strict criterion regarding appointment of independent directors and it should be mandatory for firms to take approval from SECP before appointment of independent directors.It is proposed that SECP and PICG should launch awareness programs and highlight potential benefits of CG compliance.Political interference is quite common among business organizations in Pakistan; hence, politics should be separated from business and this can only be done through enforcement.Corruption is another severe problem that exist in Pakistan.Government needs to take reforms to tackle corruption in the country and strict action is proposed against those who involved in it.Social barriers are another big challenge in corporate sector of Pakistan.
Emerging markets have institutional differences from developed markets, and these should be integrated to CG policies while adapting the CG codes from developed markets.In addition, the policies designed for developed markets may be ineffective in emerging markets[12]due argued that different stakeholders may adapt socially legitimate and institutionally accessible CG practices to build different coalitions.The firms have high ownership concentration in emerging markets[26]and state or a family holds most of the stake[27].Jackson [28]argued that different forms of agency conflicts take place across different

Table 1 ) : Table 1 Cronbach Alpha Value of Instrument
Note: Done by author based on the information (survey data)

Table 3
In EFA, the next step is the extraction of factors.Researchers have argued that factors are extracted till the value of variance is maximized (54) and different methods can be employed to extract factors.This study employed the principal axis factoring (PAF) method to extract the factors.This method is a preferred approach in the presence of a multivariate normality