The Effect of Controlling Shareholders and Corporate Governance on Audit Quality

This research aims to investigate the effect of controlling shareholders and the effectiveness of board of commissioners and audit committee on the audit quality measured by AQMS (Audit Quality Metric Score). This study uses companies listed in the Indonesian Stock Exchange as the sample. The results of this research provide evidence that alignment and entrenchment effect of the controlling shareholders have positive effect on audit quality. The alignment of interests between the controlling and non-controlling shareholders cause the company to appoint a high-quality auditor. The results also show that when companies face high entrenchment effect of the controlling shareholders, they still appoint a high-quality auditor to reduce the agency conflict and to maintain the company's reputation. This research also shows that the effectiveness of board of commissioners and audit committee positively affect the audit quality.---Penelitian ini bertujuan untuk menguji pengaruh pemegang saham pengendali serta efektivitas dewan komisaris dan komite audit terhadap kualitas audit yang diukur dengan AQMS (Audit Quality Metric Score). Penelitian ini menggunakan Perusahaan yang terdaftar di Bursa Efek Indonesia sebagai sampel. Hasil penelitian ini memberikan bukti bahwa efek alignment dan entrenchment dari pemegang saham pengendali berpengaruh positif terhadap kualitas audit. Penyelarasan kepentingan antara pemegang saham pengendali dan pemegang saham non-pengendali menyebabkan Perusahaan menunjuk auditor yang berkualitas tinggi. Hasil penelitian juga menunjukkan bahwa ketika Perusahaan menghadapi efek entrenchment yang tinggi dari pemegang saham pengendali, Perusahaan tetap menunjuk auditor yang berkualitas tinggi untuk mengurangi konflik agensi dan untuk mempertahankan reputasi Perusahaan. Penelitian ini juga menunjukkan bahwa efektivitas dewan komisaris dan komite audit berpengaruh positif terhadap kualitas audit.


INTRODUCTION
explained that agency problem between management and shareholders commonly arises in companies with dispersed ownership structure. The dispersed ownership on the hands of many shareholders discourages the shareholders to monitor performance and decision making of management so that the control of the company lies on the management's hands (Coffee 2010). Companies with disperse ownership structure can usually be found in the common law countries with strong investor rights protection, such as the United States, the United Kingdom, and Canada (LaPorta et al. 1999).
Unlike in the common law countries, in civil law countries with weak investor rights protection, the majority of companies tend to have ownership concentrated in the hands of a few shareholders. This is proven by Claessens et al. (2000) through their research on 2,980 companies in Asia, including 132 Indonesian companies, which the results show that the ownership of public companies in Asia tends to concentrate in a family ownership.  stated that when ownership structure is concentrated in a few shareholders, controlling shareholders would have the ability to determine the company's direction and operation, which is commonly referred as entrenchment effect . The presence of the entrenchment effect, however, does not always result in an agency conflict in the company.
The agency conflict between controlling and non-controlling shareholders occurs when the controlling shareholders, with the control they exert, use the discretion of the company according to their personal interests, and therefore could potentially harm the interests of the non-controlling shareholders . This phenomenon is also called negative entrenchment effect .
The negative entrenchment effect potentially worsened when the company is controlled by ultimate controlling shareholders through a pyramidal ownership mechanism 1 . The pyramidal ownership mechanism enables the ultimate controlling shareholders to have the control rights 2 far exceeding their cash flow rights 3 . This may motivate the ultimate controlling shareholders to expropriate the wealth of the company without bearing any high cost if there is a financial loss or a decline of the company's value (Claessens et al. 2000).
In the condition when control rights exceed cash flow rights, the ultimate controlling shareholders may be motivated to appoint a public accounting firm (PAF) with low audit quality in order to maintain the condition of asymmetric information with external parties, so that the chance that their expropriation is discovered becomes lower. Choi et al. (2007) proved in their research that the greater the difference between control rights and cash flow rights of the ultimate controlling shareholders, the higher the probability that the company would appoint a PAF with lower audit quality.
On the other hand, the ultimate controlling shareholders may also be motivated to appoint a PAF with higher audit quality in order to raise the investors' trust towards the quality of financial statements. Fan and Wong (2005) show that the higher difference between control rights and cash flow rights of the ultimate controlling shareholders in a company, the higher the probability the company chooses a PAF with higher audit quality.
Furthermore, as the percentage of ownership of the controlling shareholders increases, the entrenchment effect will be reduced and replaced with alignment effect (Fan and Wong 2002). The decrease of the expropriation ability was due to the higher costs incurred by the controlling shareholders if the company suffers a loss or declining share 2 Controlling right is voting right to participate in determining important discretions in the company. 3 Cash flow right is financial claim of the shareholders to the company. valuation . The increase in ownership of the controlling shareholders in turn could trigger the controlling shareholders to increase the value of the company and align their interests with non-controlling shareholders. This effect is called alignment effect (Fan and Wong 2002). The alignment effect of controlling shareholders would motivate controlling shareholders to improve the quality of audit of financial statements that are expected to increase the value of investor confidence to the quality of the company's financial statements.
Motivation of ultimate controlling shareholders in choosing PAF may be affected by the role of Board of Commissioner (BOC) and audit committee. According to OECD (2004), the BOC has a role in "ensuring the integrity of the corporation's accounting and financial reporting systems, including the independent audit, and that appropriate systems of control are in place, in particular, systems for risk management, financial and operational control, and compliance with the law and relevant standards". Lin and Liu (2009) found that companies with appropriate size of BOC have positive effect on the selection among of the Top 10 auditors, because more members in the BOC improve the monitoring effectiveness. It can be concluded that the role of BOC has positive effect on audit quality.
In addition, the audit committee also has a role in the auditor selection process. One of the audit committee's roles is to recommend BOC regarding the appointment of a PAF based on considerations of independence, assignment scope, and audit costs (Regulation of the Bapepam-LK (now known as OJK (Otoritas Jasa Keuangan) or Financial Services Authority) Number KEP-643/BL/2012). Thus, it can be seen that the audit committee has the ability to improve the audit quality of the company's financial statements. Such ability depends on the effectiveness of the audit committee itself. Abbott and Parker (2000) show that the higher the level of audit committee's effectiveness, the higher the tendency of the company to select a PAF with higher audit quality. This research is important to be conducted in Indonesia because more than 50% of companies in Indonesia have concentrated ownership structure (Claessens et al. 2000;Diyanty 2012). The pyramidal ownership mechanism enables the controlling shareholders to have control rights exceeding their cash flow rights. This research contributes the literature in several ways. First, this research tests the effect of ultimate family controlling shareholders on the appointment of auditor in the Indonesian context which has a high family concentrated ownership structure. Previous research only accounts for blockholders ownership, while we trace the ownership until the biggest ultimate controlling ownership of the companies. Second, this research tests the effect of the role of Board of Commissioners (BOC) and audit committee on the appointment of auditor, using a comprehensive measurement from Hermawan (2009). Third, this research uses more comprehensive measurement of audit quality that consist of several measurements which are the size of the PAF (Big 4 or non-Big 4), the audit tenure, the auditor industry specialization, and the perspective of independence which is measured by the importance of the client to the auditor (client importance) and the availability and the accuracy of the going concern audit opinion (a proxy for audit failure). More comprehensive measurement of audit quality enables this research to test the effect of ultimate controlling shareholders and corporate governance mechanism on the audit quality in a more robust testing.
According to above background, this research aims to investigate the effect of controlling shareholders through entrenchment and alignment effects on the audit quality measured by Audit Quality Metric Score which consists of auditor size, the audit tenure, the auditor industry specialization, client importance, and the going concern audit opinion. This research also aims to investigate the effect of the role of board of commissioners (BOC) and audit committee on audit quality. This research expands previous researches by measuring audit quality by the degree of competence, industry specialization and audit tenure, and also, independence (Herusetya 2012). To examine the alignment and entrenchment effect of ultimate controlling shareholders, this research will trace the chain of companies' ownerships to the ultimate controlling shareholders. The ultimate controlling shareholders are shareholders, individual or family group, government or foreign companies, with the highest control rights at the chain of company's ownership.
The tracing method of the ownership chain refer to research by Diyanty (2012), if there are more than one controlling shareholders from the same family, the total ownership would be the total ownership of the family.
The measurement of audit quality is based on the degree of competence, industry specialization and audit tenure, and independence (Herusetya 2012). The measurement of the role of BOC and audit committee is based on independence, competence, audit activity and the size of the audit committee (Hermawan 2009).

Ownership Structure and Agency Conflict
In a dispersed ownership structure, agency conflict commonly occurs as a consequence of management's discretions that are not in accordance with shareholders' interests (Jensen and Meckling 1976). In concentrated ownership structure, agency conflict commonly occurred between controlling shareholders and management with non-controlling shareholders. This is because the controlling shareholders may control the management, so that management's discretions are frequently used to gain the benefits for the controlling shareholders and neglect the rights of the non-controlling shareholders (Fama and Jensen 1983;LaPorta et al. 1999;Claessens et al. 2000;Diyanty 2012). With higher control rights, the controlling shareholders could control discretions, both in direction and strategic decision making of the company. This condition is also known as entrenchment effect (Coffee 2010).
Entrenchment effect potentially worsens in the companies that have pyramidal ownership structure. Pyramidal ownership structure allows controlling shareholders to have control rights higher than cash flow rights. Accordingly, controlling shareholders could freely undertake activities regardless the non-controlling shareholders' interests without bearing any high costs in the event of loss or declining company's value (Claessens et al. 2000).
On the other hand, according to Fan and Wong (2002), the increase in the ownership could reduce the expropriation ability of the controlling shareholders and also encourage them to increase the company's value, which is called as the alignment effect. This is because controlling shareholders bear the risk bear a greater risk for the failure of the company (Fan and Wong 2002). The higher risks would lead the controlling shareholders lower the expropriation actions and maintain the company's credibility by increasing management monitoring .

Audit Quality
Audit quality is a complex and multidimensional concept (Herusetya 2012). The perception of audit quality is different among stakeholders of the company, depends on their involvement level in audit process and their point of view on how they measure audit quality.
For example, investors have their own perspective towards audit. Investors want that the financial statements they use are useful for decision making. To be useful for decision making, financial statements should have high credibility, so investors measure audit quality from the credibility of the financial statements. Investors would review the preparer of the financial statements and auditor who has given an opinion to the financial statements. Investors may expect companies audited by auditors with good reputation to produce credible financial statements.
The early concept of audit quality is traced back to the auditors' independence.
According to DeAngelo (1981), auditors' independence is defined as the probability that auditor would find and report misstatements on financial statements and would not mind the pressure from management to not report the misstatement (if there is pressure from management).
Audit quality could also be viewed from the perspective of audit failure. Audit failure is difficult to describe, nevertheless according to Francis (2004) it could be measured from various sources, such as litigations to the auditor, business failure, examination by the stock market authority, and the restatement of financial statements. The higher the audit quality, the lower the audit failure would be.
Auditor size may also be employed as a measure of audit quality. DeAngelo (1981) argues that large public accounting firms (PAFs) have better audit quality due to lack of dependence on certain clients, so they have greater independence. Inspired from that study, studies focused on Big 4 PAFs contended that Big 4 PAFs have had reputation and incentive to provide high quality audit service to maintain their reputation (Simunic and Stein 1987;Francis and Wilson 1988). Becker et al. (1998) and Francis et al. (1999) demonstrate that companies audited by Big 4 PAFs have lower abnormal accruals denoting lower earnings management and higher earnings quality.
The Big 4 audit quality, under several studies, may arise from higher audit fees and special expertise in the industry. Simunic (1980) discovers that Big 4 PAFs have higher audit fees (fee premium) than other PAFs after controlling client characteristics i.e. size, complexity, and risk sharing between auditors and clients. On average, Big 4 audit fees are estimated to be 20% higher than non-Big 4. Higher audit fees might improve audit quality as greater audit effort, shown by lengthier audit working hours or more competent auditor (Francis 2004).
If a PAF has a lot of clients in a particular industry, it should have opportunity to enhance its ability and gain experience until the PAF becomes expert in that industry. Big 4 PAFs have plenty clients across industries and resources to improve the abilities of their auditors, so that the Big 4 is more likely to develop an industry expertise compared to non-Big 4 PAFs. Balsam et al. (2003) detected that clients audited by Big 4 PAFs having industry specialization had lower discretionary accruals and higher earnings response coefficient describing higher earnings quality.
Audit tenure, or how long a PAF auditing a client, may also be used to measure audit quality. Theoretically, at the beginning of the tenure, audit quality would be low since the PAF is still in the process of understanding the client's business. Johnson et al. (2002) found the evidence that there was lower audit quality in the first three years after auditor switch. However, too long tenure may have adverse effect on audit quality because relationship between the auditor and the client would be closer, so that the independence and the professional skepticism of the auditor would be reduced (Johnson et al. 2002).
Financial statements users make economic decisions based on audited financial statements. In consequence, the opinion of the company's ability to continue its business is extremely important to financial statements users. Going concern opinion, clearly stating the auditors' doubt of the company's ability to continue its business, is a signal that the company is facing going concern problems, such as financial problems. Therefore, going concern opinion could be a measure of audit quality. It is considerably necessary for auditors to provide accurate going concern opinion because according to Francis (2004), false positives in going concern opinion (company obtaining going concern opinion but not going bankrupt) may reduce audit quality. The empirical research of Bhimani et al. (2009) verified that generally companies receiving going concern opinion indeed went bankrupt, and the probability that such companies going bankrupt is greater than companies not receiving going concern opinion. Lee and Stone (1995) concluded that components of auditor quality can be summarized as competence (skills, knowledge and experience) and independence (lack of prejudice). Schandl (1978) and Flint (1988) regarded that independence is more important to an auditor than competence, but both qualities are required for an effective audit. On the other hand, Watts and Zimmerman (1986) stated that the issue of independence and competence in the case of auditor reporting a contractual breach is conditional on two separate probabilities. Moizer (1991) offered an alternative to the two views, and suggested that competence and independence of an auditor is a choice: a competent auditor may choose to be dependent or independent. Departing from Moizer's (1991) view, Lee and Stone (1995) proposed that auditor cannot choose to be independent unless he is competent. Using case study from several court decisions (e.g. Westminster Road Construction & Engineering Co. Ltd. (1932,86 Acct. L.R. 38) and Thomas Gerrard and Son Ltd. (1968, Ch. 455)), they argued that because the auditors presented in the cases were not competent, they had to depend on evidential materials and figures provided by the clients' management, and unable to question them in an independent manner.
In the studies mentioned above, it can be concluded that audit quality could be measured in many dimensions so that it could entirely describe audit quality (Bamber and Bamber 2009;Francis 2004;Watkins et al. 2004). Darmadi (2012) found that concentrated shares ownership, measured by the percentage of shares owned by the largest shareholders, has positive effect on audit quality. While Fan and Wong (2002) show that the increasing ownership of the controlling shareholders may reduce the expropriation ability of the controlling shareholders. The increase in controlling shareholders' ownership increases the alignment effect, where such increase would encourage the controlling shareholders to increase the value of the company.

Effect of Alignment Effect of Controlling Shareholders on Audit Quality
Based on above argument, the alignment effect of the controlling shareholders is expected to enhance audit quality. H1: The alignment effect of the controlling shareholders positively affects audit quality.

Effect of Entrenchment Effect of Controlling Shareholders on Audit Quality
The entrenchment effect of the controlling shareholders is the ability of the controlling shareholders to direct discretions of the company . The pyramidal ownership mechanism enables the controlling shareholders to have control rights exceeding cash flow rights. This condition encourages expropriation of the company's wealth without bearing any high cost when there is loss or decrease in the company's value, because the controlling shareholders have low cash flow rights .
When the expropriation activity is detected by external parties, for instance investors, investors would discount their valuation of the company's value, which may cause the decrease of shares value and the increase of cost of capital Fan and Wong 2005). In such situation, the controlling shareholders may be encouraged to embezzle their expropriation activities so that the external parties could not detect them by decreasing the disclosure quality of the company's financial statements (Fan and Wong 2002). To help concealing the real financial conditions of the company, the company may appoint auditor with low audit quality (Choi et al. 2007(Choi et al. , 2008.
Based on the above researches, one of the probabilities that may arise from the entrenchment effect of the controlling shareholders is a low audit quality. This low audit quality is caused by the desire of the controlling shareholders to conceal their expropriation activities by reducing the transparency of the financial statements and this could potentially lower the audit quality.
However, agency conflict is not always negatively associated with audit quality (Fan and Wong 2005;El Ghoul et al. 2007). The presence of agency conflict may reduce company's value and increase the cost of capital, also complicate the controlling shareholders to obtain outside funding (El Ghoul et al. 2007). In that condition, the controlling shareholders would be motivated to appoint high-quality external auditor to reduce the potential agency conflict caused by the negative entrenchment effect (Fan and Wong 2005). Additionally, the appointment of high-quality external auditor may also be viewed as a signal that the controlling shareholders would protect and concern about the interests of the non-controlling shareholders. Because of the contradicting view about the effect of controlling shareholders on the audit quality, therefore, the next hypothesis can be stated as two-tail hypothesis as follows: H2: The entrenchment effect of the controlling shareholders affects the audit quality.

Effect of Board of Commissioners and Audit Committee on Audit Quality
In two-tier corporate governance system, the role of the board of commissioners (BOC) is to conduct monitoring function to promote accountability and transparency of the presentation of financial statements (National Committee of Indonesia Governance Policies 2006; Panel 1994 in Abbott and Parker 2000). In carrying out their duties, BOC is assisted by audit committee.
According to the Regulation of the Bapepam-LK (now known as OJK (Otoritas Jasa Keuangan) or Financial Services Authority) Number KEP-643/BL/2012, audit committee is a committee established and is responsible to BOC in assisting them to do their duties and responsibilities. The audit committee is established by BOC to promote accountability and transparency of the presentation of the company's financial statements so that the company may mitigate risks of reputational and financial loss (Menon and Williams 1994). One of the main duties of the audit committee according to the Regulation of the Bapepam-LK Number KEP-643/BL/2012 is to give recommendation to BOC regarding the appointment of PAF which is based on independence, engagement scope and fee of the PAF. Therefore, the audit quality of a company is greatly influenced by the role of BOC and the audit committee. Maharani (2012) found that the size of the board of commissioners has a positive and significant effect to the appointment of auditors of great quality. According to OECD (2004), the board of commissioners affects the monitoring capacity of management's conduct, and minimize information the information asymmetry between the management and the owners by increasing the transparency of financial reporting. Putra et al. (2014) found that the independence of the board of commissioners affects the choice of quality auditors. Lin and Liu (2009) found that companies which have large supervisory board (the board contains more members) has a positive effect to the appointment of Top 10 auditors. Beasley and Petroni (2001) found that the independence of the board is related with choosing of auditors who has industry specialization. In general, it can be concluded that the board of commissioners has a positive effect to audit quality.
The quality of audit committee is also considered as an important factor that influences audit quality (Fitriany 2011). Rustam et al. (2013) found that the activeness and independence of the audit committee has a significant and positive effect to the audit fees as a measure of audit quality. The effectiveness of audit committee is also found by Maharani (2012) to have a positive and significant effect to the appointment of high quality auditor. The appointment of high-quality auditor is caused by the aspiration of the audit committee to conduct their monitoring function of the financial statements effectively (Wild 1996;Abbott and Parker 2000). It can be concluded from the previous research that the audit committee also has a positive relationship with audit quality.
According to previous researches above, the effective role of the board of commissioners and audit committee is expected to increase audit quality. Therefore, the next hypothesis can be stated as follows: H3: The effectiveness of the board of commissioners and audit committee positively affect the audit quality.

Research Model
The tests about the alignment and entrenchment effect of the controlling shareholders, and the effect of the effectiveness of the board of commissioners and the audit committee will be conducted using a proportional odds model that is processed using ordered logistic method. The ordered logistic method is used because the dependent variable is an ordinal variable.
The model specification to test hypotheses H1, H2, and H3 is as follows:

Variables Operational Definition
The definition and measurement of the variables in this research can be seen in Table  1. Explanation of the variables will follow after the table. The cash flow rights is the addition of the multiplication of the percentage of share ownerships for every chain of share ownership.
Diyanty (2012) 2 Cash flow leverage (CFL) Cash flow leverage is the ratio between control rights (CR) and cash flow rights (CFR).
CRit is the value of control rights is computed using the weakest link of the control chain. However, if there is more than one individual in a family, their ownership proportion will be combined to one and then the weakest link is examined. The score of the effectiveness of the board of commissioners and the audit committee (GOV) The score is measured using a questionnaire divided to parts that measure the effectiveness of the board commissioners (17 questions) and the effectiveness of the audit committee (11 questions). Every question can have value between 1 and 3, therefore the score of the effectiveness of the board of commissioners and audit committee has a minimum score of 28 ((17 x 1) + (11 x 1)) and a maximum score of 84 ((17 x 3) + (11 x 3)).
Hermawan (2009)   4 Leverage of the company (LEV) Leverage is measured by dividing the total of long-term debt to total assets. − Long-term debtit = long-term debt of company i in year t Total assetsit = total assets of company i in year t Grossman and Hart (1982) 5 Size of the company (SIZE) The size of the company is measured using the natural logarithm of market capitalization of the company in the end of year t. 7 The level of audit quality (AUDQUAL1) The audit quality in this research is measured using AQMS (Audit Quality Metric Score) formulated by Herusetya (2012). AQMS is measured by computing the score of 5 measures of audit quality from the perspective of auditor competence; that is the size of the PAF (Big 4 or non-Big 4), the audit tenure and the auditor industry specialization; and the perspective of independence, which is measured by the importance of the client to the auditor (client importance) and the availability and the accuracy of the going concern audit opinion (a proxy for audit failure). The maximum value of every audit quality measures is 1, so that the maximum value of audit quality measured by AQMS is 5.

No
Audit Quality Measure Description 1 PAF Size Valued 1 if the PAF is one of the Big 4, and 0 otherwise.

Industry Specialization
Valued 1 if the PAF has the greatest share in the industry and 0 otherwise. The greatest industry share is measured with the threshold of highest 10% of industry share (Craswell et al. 1995). The industry share is measured with the ratio: The total of a PAF clients' assets in the manufacturing industry The total assets of all PAF clients' in the manufacturing industry 3 Audit Tenure Valued 1 if the tenure is 3-4 years, and 0 otherwise 4 4 Client Importance (CI) Measures the economic dependence level of the auditor to the client. It is measured with the ratio:

= ∑ =1
Where the numerator is the natural logarithm is the total assets of client i on year t and the denominator is the natural logarithm of the total assets of clients audited by PAF i in year t 5 . If CI is valued between ± , then this variable is valued 1, and 0 otherwise. 5 The Accuracy of Going Concern Opinion (GC) Valued 1 if: a. The PAF issued going concern opinion to client i on year t and at year t + 1 that client i experienced negative cash flows from operations or net loss; or b. The PAF did not issue a going concern opinion to client i on year t and on year t + 1 that client i did not experience negative cash flows or net loss. And valued 0 otherwise. AQMS Maximum Value = 5 Herusetya (2012) 4 According to previous research (Herusetya 2012), auditors will obtain a reasonable understanding of their clients' business and industry when audit tenure is 4-8 years without lowering auditors' independence. See also Johnson et al. (2002), where audit tenure is grouped into 2-3 years (short tenure), 4-8 years (medium tenure), and over 9 years (long tenure). However, in this research, a regulation (PMK No. 17 Year 2008) that limits audit tenure to 5 years is in effect. Therefore, we find it reasonable that 3-4 years is the 'medium tenure' when the audit tenure is limited to 5 years. 5 The data for auditor and clients are obtained from PPAJP (Pusat Pembinaan Akuntan dan Jasa Penilai) of the Ministry of Finance. After 2014, it is called Pusat Pembinaan Profesi Keuangan/PPPK. The data is then connected to asset data from the financial statements of companies.

The Alignment and Entrenchment Effect of the Controlling Shareholders
In this research, the controlling ownership of the controlling shareholders is measured using two variables, the cash flow rights/CFR (for the alignment effect), and the ratio between the control rights and cash flow rights/CFL (for the entrenchment effect). The measuring of variables is based on Diyanty  Claessens et al. (1999aClaessens et al. ( , 2000;  which is conducted by tracing to the ultimate owners of the company. If the ultimate owners amounted to more than one individual in one family, the total ownership is the total ownership of the family. The data for family ownership is obtained from Diyanty (2012) from Pusat Data Bisnis Indonesia

Cash Flow Right (CFR)
This variable describes the cash flow rights of the greatest controlling shareholders (Diyanty 2012). The cash flow rights is the addition of the multiplication of the percentage of share ownerships for every chain of share ownership.

Control Rights (CR)
The value of control rights is computed using the weakest link of the control chain (LaPorta et al. 1999;LaPorta et al. 2002;Claessens et al. 1999bClaessens et al. 2000). However, if there is more than one individual in a family, their ownership proportion will be combined to one and then the weakest link is examined (based on Diyanty's (2012) method).

Cash Flow Leverage (Ratio between Control Rights and Cash Flow Rights/CFL)
According to LaPorta et al. (1999), the high ratio between the control rights and cash flow rights happened when the controlling shareholders reduced their ownership through superior voting rights through a pyramidal structure or cross-ownership. Cash flow leverage is a ratio that measures the incentive of expropriation of the controlling shareholders and the entrenchment effect of the controlling shareholders (Diyanty 2012).

The Role of the Board of Commissioners and Audit Committee (GOV)
The role of the board of commissioners and audit committee (GOV) is measured with an index developed by Hermawan (2009) related to the size, independence, competence, and the activity of the board of commissioners and the audit committee. This index is divided to parts that measure the effectiveness of the board commissioners (17 questions) and the effectiveness of the audit committee (11 questions). Every question can have value between 1 and 3, therefore the score of the effectiveness of the board of commissioners and audit committee has a minimum score of 28 ((17 x 1) + (11 x 1)) and a maximum score of 84 ((17 x 3) + (11 x 3)). The data for the index is obtained from the Annual Report of the companies.

Company's Size (SIZE)
Previous studies found that the appointment of PAFs tend to have a positive relationship with the size and business complexity of the company (Beatty 1993;Fan and Wong 2005). The size of the company is measured using the natural logarithm of market capitalization of the company in the end of year t. Willenborg (1999) found that companies audited by big-scale auditors have higher profitability, ceteris paribus. Companies that have higher profitability also tend to appoint auditors with bigger scale (that exhibit higher audit quality) (Chaney et al. 2004). Profitability of the company is measured using Return on Assets (ROA). The ROA is computed by dividing net income with average of total assets.

Company's Leverage (LEV)
Companies with greater leverage tend to have higher bankruptcy risk or financial failures (Grossman and Hart 1982). According to Grossman and Hart (1982) also, in that situation, companies tend to appoint a betterquality auditor to avoid the decrease in the company's value. Leverage is measured by dividing the total of long-term debt to total assets.

Audit Quality
According to the Ministry of Finance (MoF) Regulation No. 17 Year 2008, a company should rotate PAFs every 6 book years and PAF partner every 3 years. A change of 50% or more of the partner's name counts as a change in PAF. Fitriany (2011) found that many non-Big 4 PAFs merged themselves to circumvent the rotation requirements, where the operational of the PAF still rested within the previous PAFs and only the name changed. 6 To address this issue, Fitriany (2011) divided the rotation of PAFs into real and pseudo rotation, therefore in this research tenure is also divided into real and pseudo tenure. Pseudo tenure is defined as tenure measured to five years before the audit engagement in the research period (if the information is available), whether there is a partner change in the PAF in the five years period before the audit engagement. The real tenure is measured without regarding the change of partners. If the PAF still has the same affiliation, it will be counted as one PAF.
As an illustration, Company A is audited by PAF Purwantono, Sarwoko, and Sandjaja affiliated to Ernst and Young (EY) for the period 2004-2007. For the period of 2008 to 2010, the company is audited by PAF Purwantono, Suherman and Surja affiliated to EY. If measured using pseudo tenure, at the end of year 2008, the tenure of the PAF is 1 years (because it is currently audited by PAF Purwantono, Suherman, and Surja). However, if measured using real tenure, Company A has been audited by the PAF for 5 years, because the PAF is still affiliated with EY for 5 years before 2008, so that it is counted as one PAF. The time limit of 5 years is used because the data for PAF tenure from the PPAJP is obtained from 2004. Fitriany (2011) found that in the period after the enactment of the Ministry of Finance (MoF) Regulation) No. 17 Year 2008, the rotation of PAF partner increased the audit quality from the perspective of neutrality and predictability. Fitriany (2011) also found that in that period the audit quality from the perspective of neutrality has a convex-shaped relationship with the audit tenure and from the perspective of predictability has a linear negative relationship. In this research, we use sample from 2008-2012, and according to MOF Regulation No. 17 Year 2008, the rotation of PAF partner must be conducted every 3 years 7 . Based on the research and the regulation, this research uses 3-4 years as tenure that is considered "good" because every 3 years there must be a rotation in the PAF partner, but the audit quality will deteriorate when the tenure is too long. Therefore, tenure is considered good if it lasts for 3 to 4 years.
The data for modified AQMS variable is obtained from the financial statements of the companies, Indonesia Capital Market Directory, and the Pusat Data Bisnis Indonesia (PDBI).

Population and Sample
This research uses secondary data from manufacturing companies listed in the Indonesia Stock Exchange (IDX) from 2008 to 2012. Manufacturing sector is the sector with the most companies in IDX, so we expect it to be representative to the whole market. We also consider the time and cost needed to trace the ultimate owners and calculate the audit quality. The data is obtained from Annual Reports of companies from the IDX and Data stream Thomson Reuters and the tracing of ownership structure is obtained from the Ministry of Law and Human Rights of the Republic of Indonesia. The data about PAF is obtained from the Pusat Pembinaan Akuntan dan Jasa 7 Indonesia Government Regulation No. 20/2015 has lifted the restriction that PAFs have to be rotated every 5 years. However, when this study is conducted (2014) The population in this research is all companies listed in the IDX from 2008 to 2012. The samples are chosen using purposive sampling, a sample choosing method according to certain criteria, that is: manufacturing companies listed in the IDX from December 31, 2008 until December 31, 2012 (that never delisted, suspended, or went private), companies whose share ownership can be traced to the ultimate shareholders and the entrenchment effect can be measured, companies with positive equity 8 , companies that are not Foreign But Indonesian (FOBINDO) 9 , and companies that have all components that is required to measure the variables used in this research. According to those criteria, the samples used in this research are 432 companies for 5 years (2008 until 2012).

Descriptive Statistics
The descriptive statistics of the data is shown on Table 3. From Table 3, it can be seen that the audit quality has the average of 2.8866. This value shows that on average, the audit 8 Companies with negative equity generally experienced cumulative loss continually and tend to use debts to finance their operations. This condition can affect the basic assumption of the creation of the company's financial statement, the going concern assumption (IFRS conceptual framework). 9 FOBINDO are companies that when established were owned by families, but then changed ownership to foreign companies in the next years, with control still maintained in the establishing family (Kim 2003in Diyanty 2012. FOBINDO is identified by Diyanty (2012) that was collected from Pusat Data Bisnis Indonesia (PDBI). For example, Diyanty (2012) gave example of Indoofod that, before the 1997-1998 Indonesian economic crisis, was owned by Salim group. In 1999, Indofood was acquired by First Pacific and Nissin (foreign direct investment companies). However, First Pacific, which is situated in Hong Kong, quality is moderate (from the max score 5). The descriptive statistics shows that CFL on average has a value above 1, which means that in most companies, the controlling shareholders have control rights exceeding their cash flow rights. The effectiveness of the board of commissioners and audit committee is actually owned by the Salim group (Kompas, January 26, 1999in Diyanty 2012. Diyanty (2012) suspects that the previous Indonesian company controllers may have purchased shares using foreign companies to control the companies they previously owned. We determine if companies are still controlled by the same owners after 1999, even if the name of the controllers change, based on the information from PDBI. If there is a change in controllers of a company, but it is still actually owned by the previous owners, it will be identified as FOBINDO and excluded from our sample. 10 This condition is distinct from "financial statements not found". In some cases, the financial statements is found, but the page containing the audit opinion is not found, or financial informations are found from other sources such as Indonesia Capital Market Directory, but the financial statements itself is not found.
has an average score of 0.68 (maximum value of 1). It is consistent with World Bank (2010) that states that there are some weaknesses in the board of commissioners and audit committee in the monitoring function in Indonesia.

The Influence of the Alignment Effect of the Controlling Shareholders to Audit Quality
The aim of this research is to test the influence of the alignment effect of the controlling shareholders and the effectiveness of the board of commissioners and audit committee to the audit quality. The result of the hypothesis testing can be found in Table 4.
Based on Table 4, the result shows that if CFR increases, the company tends to choose a higher quality auditor. This result supports the results of the researches of Darmadi (2012) andHay et al. (2008). A high alignment effect causes the controlling shareholders to have a low motivation to expropriate the noncontrolling shareholders (Diyanty 2012). Diyanty (2012) also stated the increase of the share ownership will increase the alignment of the interest of the controlling and the noncontrolling shareholders. The alignment of interest between the controlling and the noncontrolling shareholders is the factor that encouraged the company to appoint a highquality auditor.
According to Table 4, the result of CFL variable showed that the probability of higher quality audit level compared to lower audit quality level will increase by 2.2898 times if the CFL increases by 1 time. This result implies that as the entrenchment effect of the controlling shareholders is stronger, the company tends to choose a high-quality auditor.
The finding of the entrenchment effect in this research supports the result of Fan and Wong (2005) that stated that the controlling shareholders will choose a high-quality auditor to give a signal to the non-controlling shareholders that they care for the interests of the non-controlling shareholders. Additionally, the controlling shareholders wanted to maintain the reputation of their company by appointing a high-quality auditor (El Ghoul et al. 2007).
The effectiveness of BOC and Audit Committee or GOV gives result as predicted, with a positive and significant coefficient (alpha 1%). This result shows that BOC and Audit Committee have a positive and significant effect to the quality of auditor appointed by the company. The odds ratio value of 6.3240 shows that in every 1-point increase of GOV, the probability of audit quality that is higher compared to lower audit quality will increase by 6.3240 times. The existence of an effective BOC and audit committee is able to strengthen the motivation for controlling shareholder to appoint a public accounting firm with higher audit quality.
This result is consistent with Lin and Liu (2009) that found that the size of the supervisory board (the board of commissioners) has a positive effect in the appointment of Top 10 auditor, and Beasley and Petroni (2001) that found that the independence of the board is associated with the appointment of a higher quality auditor. The result of this research is also in line with the findings of Rustam et al. (2013) that the activity and the independence of the audit committee have a positive and significant relationship on audit quality.
The significant result of PROF variable shows that the probability of a higher quality audit compared to a lower quality audit will increase by 8.6467 times with 1-point increase in PROF. The interpretation of the odds ratio for the SIZE variable is that in every 1-point increase in SIZE, the probability of a higher quality audit compared to a lower quality audit will increase by e 1,2989 or 3.6653 times. These findings are consistent with Beatty (1993), Fan and Wong (2005), and Lennox (2005).

Sensitivity Analysis
To prove that the research model will give a consistent result if the proxy of the variable is changed, a sensitivity analysis is conducted. The sensitivity analysis conducted in this research is to change the tenure in the AQMS to a real tenure (so that the variable AUDQUAL1 becomes AUDQUAL2). This change in tenure is conducted because there is a probability that the rotation of the PAF done by the companies is not a real but rather only a pseudo rotation (Fitriany 2011).
In this sensitivity analysis, like the main analysis, a test about the alignment and entrenchment effect of the controlling 11 In an ordered logistic regression, Stata determines constant to be 0, and estimate cutpoints (threshold) to separate levels of the response (Y) variable. _cut is the intercept of the Y variable when X = 0. For example, _cut1 is the area where Y = 0, _cut2 is the area where 0 < Y ≤ 1, _cut3 is the area where 1 < Y ≤ 2, and so on.
shareholders are also conducted. The aim is to find out if the change in tenure in AQMS gives a robust result.
From the result, it can be seen that the coefficient of CFR is positive and significant. This result is consistent with the main analysis. It can be concluded that if CFR increases, the company will tend to appoint a higher quality auditor. The high alignment effect of the controlling shareholders results in the low motivation of the controlling shareholders to conduct expropriation (Diyanty 2012). The presence of alignment of interest between the controlling and non-controlling shareholders becomes a motivation for the company to appoint a high-quality auditor.
The CFL variable is significant, and this result is consistent and supportive of the main analysis. This result implies that if the entrenchment effect of the controlling shareholders is stronger, the company will tend to choose a higher quality auditor. The finding about entrenchment effect in this research supports the research of Fan and Wong (2005) that stated that controlling shareholders will choose a high-quality auditor to give signal to the non-controlling shareholders that they care about the interests of the non-controlling shareholders. According to El Ghoul et al. (2007), the controlling shareholders also tend to appoint high quality auditors to maintain the reputation of their company and to avoid litigations.
The effectiveness of BOC and Audit Committee or GOV also shows a positive and significant effect to the quality of auditor appointed by the company. Every result of this sensitivity analysis supports the main analysis. The conclusion is that the sensitivity analysis in this research gives a robust result.

CONCLUSION
The alignment effect of the controlling shareholders has a positive effect to audit quality. The alignment effect increases the alignment of interests between the controlling and non-controlling shareholders (Diyanty 2012). The alignment of interests between the controlling and no controlling shareholders caused the company to appoint a high-quality auditor.
The entrenchment effect of the controlling shareholders turns out to have a positive effect to the audit quality. The high audit quality when the entrenchment effect of the controlling shareholder exists comes from the desire of the controlling shareholders to reduce the agency conflict by appointing a high-quality auditor (Fan and Wong 2005). The company may also appoint a high-quality auditor to maintain the reputation of the company (El Ghoul et al. 2007).
The effectiveness of BOC and Audit Committee also has a positive effect to the quality of auditors appointed. Increasing the effectiveness of BOC and the audit committee give evidence will increase the motivation of firm to select a public accounting firm with higher audit quality.
The sensitivity analysis by changing the pseudo tenure to real tenure based on Fitriany (2011) that after the enactment of the regulation of the PAF rotation, there are PAFs that rotated their partners and even merger with each other, so that the PAF looked as if it changed, when in reality the operational was still conducted by the previous PAF. This sensitivity testing shows that the use of real tenure is consistent with the main testing, both from the direction of the relationship and the significance of effect. The result of the sensitivity analysis shows that this result of this research is robust to the changing of tenure in AQMS from pseudo tenure to real tenure.
Implications of this research are: 1) Previous research only measured audit quality based on single-quality dimensions, such as PAF size and audit fee, so for the next research, the measure of audit quality in this research can be applied. A comprehensive measure of audit quality that considers competence and independence component is needed because audit quality cannot be measured by single measure; 2) This research shows that alignment and positive entrenchment effect impacts companies to choose a qualified auditor. This is because companies with family ultimate controllers want to protect their reputation by choosing high-quality external auditors. The Government should ensure to improve regulations to improve the auditor quality. For example, improving regulations to maintain the independence of external auditors and the external auditors' competence; and 3) The government should enact a regulation to improve the effectiveness of BOC and audit committee to ensure that the role BOC and audit committee to maintain the quality of external auditors. This research has several limitations: 1) This research measures audit quality using AQMS developed by Herusetya (2012). This AQMS variable is only an addition of its composing variables, so there is a possibility that it is not yet representative of the real audit quality; and 2) The data of ownership in this research is limited to the companies whose data are available in the Ministry of Law and Human Rights. Foreign companies' data are unavailable in the Ministry of Law and Human Rights so that their ownership is unknown.
For future research, audit quality can be measured using better proxy from the previous proxies, for example, by creating a weighting for audit quality measures. We concur that competence component should weigh higher than independence components. Competence is important to ensure that audit opinions are accurate. Independence is also important, because if independence is low it may lead to the collusion of auditor and management to manipulate financial statements, as evidenced in Enron, WorldCom, and Tyco cases. As stated by Lee and Stone (1995), however, we believe that auditor competence should precede independence. We suggest the weight, with the rank as follows: 1) auditor competence (industry specialization and the accuracy of going concern opinions); 2) auditor independence (tenure); and 3) auditor size (Big 4 or non-Big 4, and client importance to the auditor).